CN113835344A

CN113835344A - Control optimization method of equipment, display platform, cloud server and storage medium

Info

Publication number: CN113835344A
Application number: CN202111413935.9A
Authority: CN
Inventors: 吕自荟; 周文闻; 邱剑; 周凡珂; 谢予丛; 李绪焜
Original assignee: Alibaba Cloud Computing Ltd
Current assignee: Alibaba Cloud Computing Ltd
Priority date: 2021-11-25
Filing date: 2021-11-25
Publication date: 2021-12-24
Anticipated expiration: 2041-11-25
Also published as: WO2023093820A1; CN113835344B

Abstract

The embodiment of the application provides a control optimization method of equipment, a display platform, a cloud server and a storage medium, wherein the method comprises the following steps: determining real-time attribute information based on at least real-time environmental information of the building; predicting load prediction information of the equipment system at the future time according to the real-time attribute information, wherein the equipment system is used for adjusting the indoor environment of the building; determining a target control strategy by using an optimization control strategy model at least based on the real-time attribute information and the load prediction information, wherein the target control strategy is a control strategy in which the building indoor environment meets the expected indoor environment and the energy consumption of the equipment system meets the energy consumption requirement in a plurality of control strategies; the optimization control strategy model is obtained based on simulation model training, the simulation model is based on attribute information and a control strategy, and the building indoor environment and the energy consumption of the equipment system are obtained through simulation calculation. According to the embodiment of the application, under the condition that the building conforms to the expected indoor environment, the energy consumption of the equipment system is saved, and the energy-saving and emission-reducing effects of the building are improved.

Description

Control optimization method of equipment, display platform, cloud server and storage medium

Technical Field

The embodiment of the application relates to the technical field of building energy consumption, in particular to a control optimization method of equipment, a display platform, a cloud server and a storage medium.

Background

Buildings often rely on different types of equipment systems to achieve indoor environment adjustment of the buildings, for example, the buildings rely on equipment systems such as air conditioning systems, heating systems, ventilation systems, and the like to achieve indoor environment adjustment. Along with the implementation of the double-carbon target, building energy conservation is taken as an important link of energy conservation and emission reduction, and the method has important significance for realizing energy conservation and emission reduction control of equipment systems mainly consuming energy of buildings. Therefore, how to optimize the control of the equipment system of the building to realize energy conservation and emission reduction of the building becomes a technical problem which needs to be solved urgently by technical personnel in the field.

Disclosure of Invention

In view of this, the embodiment of the present application provides a control optimization method for a device, a display platform, a cloud server, and a storage medium, so as to optimize control of a device system of a building, and implement energy conservation and emission reduction of the building.

In order to achieve the above object, the embodiments of the present application provide the following technical solutions.

In a first aspect, an embodiment of the present application provides a method for controlling and optimizing a device, including:

determining real-time attribute information based on at least real-time environmental information of the building;

predicting load prediction information of an equipment system at a future time according to the real-time attribute information, wherein the equipment system is used for adjusting the indoor environment of the building;

determining a target control strategy by utilizing an optimization control strategy model at least based on the real-time attribute information and the load prediction information, wherein the target control strategy is a control strategy in which the building indoor environment meets the expected indoor environment and the energy consumption of the equipment system meets the energy consumption requirement in a plurality of control strategies; the optimization control strategy model is obtained based on simulation model training, and the simulation model is based on attribute information and a control strategy, and the building indoor environment and the energy consumption of the equipment system are obtained through simulation calculation.

In a second aspect, an embodiment of the present application provides a display platform, where the display platform displays an optimized control interface of an equipment system, and the optimized control interface displays a target control strategy of the equipment system, where the target control strategy includes target values of multiple control parameters of the equipment system; the optimization control interface displays the current values of the control parameters;

the target control strategy is determined by a cloud server by utilizing an optimization control strategy based on at least real-time attribute information and load prediction information of an equipment system at a future time, wherein the real-time attribute information is determined based on at least real-time environment information of a building, the load prediction information is obtained through prediction according to the real-time attribute information, and the equipment system is used for adjusting the indoor environment of the building.

In a third aspect, an embodiment of the present application provides a cloud server, including: at least one memory and at least one processor; the memory stores one or more computer-executable instructions that are invoked by the processor to perform the method of control optimization of a device as described in the first aspect above.

In a fourth aspect, embodiments of the present application provide a storage medium that may store one or more computer-executable instructions that, when executed, implement a method for control optimization of a device as described in the first aspect above.

In a fifth aspect, an embodiment of the present application provides a computer program, which when executed, implements the method for controlling and optimizing the device according to the first aspect.

According to the control optimization method for the equipment, the simulation model is not deployed on line (the simulation model carries out simulation calculation based on the operation process of a simulation equipment system), but the optimization control strategy model obtained based on the simulation model training is deployed on line, and the optimization control strategy model determines the building indoor environment and the equipment system energy consumption corresponding to different control strategies in a machine prediction mode, so that huge calculation amount caused by the fact that the simulation model is directly used on line can be avoided, and the efficiency of on-line calculation is improved. Based on the real-time attribute information, the cloud server can determine the real-time attribute information at least based on the real-time environment information of the building, and predict the load prediction information of an equipment system at the future time according to the real-time attribute information, wherein the equipment system is used for adjusting the indoor environment of the building; therefore, the cloud server can determine a target control strategy by using an optimization control strategy model at least based on the real-time attribute information and the load prediction information, wherein the target control strategy is a control strategy in which the building indoor environment meets the expected indoor environment and the energy consumption of the equipment system meets the energy consumption requirement in a plurality of control strategies. According to the embodiment of the application, the equipment system can be controlled under the condition that the building meets the expected indoor environment, so that the equipment system meets the energy consumption requirement, and the energy-saving and emission-reducing effects of the building are improved.

Therefore, the optimization control strategy model is deployed on line to replace a simulation model, so that the calculation process of determining the indoor environment of the building and the energy consumption of the equipment system corresponding to the control strategy on line is quicker, and the calculation efficiency is higher. In addition, in the embodiment of the application, the optimization of the target control strategy is performed based on the building indoor environment and the energy consumption of the equipment system in the multiple control strategies, and the control strategy is obtained by combining the multiple control parameters, so that the optimized target control strategy can consider the multiple control parameters in the whole, the optimization effect of the global control parameter is realized, and the local optimization of the local control parameter is avoided. Furthermore, when determining the building indoor environment and the energy consumption of the equipment system corresponding to the plurality of control strategies respectively, the embodiment of the application not only considers the current real-time attribute information, but also considers the load prediction information of the equipment system at the future time, so that the finally optimized target control strategy can be attached to the load condition of the equipment system at the future time, and the accurate optimization control of the equipment system is realized. In summary, the control optimization method for the device provided by the embodiment of the application can improve the on-line computing efficiency, optimize the global control parameters of the device system, and realize the accurate optimization control of the device system; according to the embodiment of the application, under the condition that the building conforms to the expected indoor environment, the energy consumption of the equipment system is saved, and the energy-saving and emission-reducing effects of the building are improved.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.

Fig. 1 is a flowchart of a control optimization method for a device according to an embodiment of the present disclosure.

Fig. 2 is an exemplary diagram of a system architecture according to an embodiment of the present application.

Fig. 3 is an exemplary diagram of a model training phase provided in an embodiment of the present application.

Fig. 4 is a flowchart of a method for training an optimal control strategy model according to an embodiment of the present disclosure.

Fig. 5A is another flowchart of a control optimization method for a device according to an embodiment of the present disclosure.

Fig. 5B is a schematic diagram of an optimization control interface provided in an embodiment of the present application.

Fig. 6 is a block diagram of a control optimization apparatus of an equipment system according to an embodiment of the present application.

Fig. 7 is a block diagram of a cloud server.

Detailed Description

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.

The equipment system is widely applied to various buildings, and the operation and maintenance (operation and maintenance) of the equipment system mainly depends on the experience of operation and maintenance personnel at present. For example, operation and maintenance personnel set control parameters of the equipment system according to experience, the control parameters of the equipment system affect the working operation of the equipment system, and the setting of the control parameters according to the experience of the operation and maintenance personnel may possibly cause unreasonable operation of the equipment system, cause that a building cannot be adjusted to an expected indoor environment condition, generate an energy waste condition of the equipment system, and further affect the energy saving and emission reduction effects of the building.

In one example, taking the equipment system as an air conditioning system, a typical water-cooled central air conditioning system may include a chiller, a cooling tower, a cooling water pump, a chilled water pump, an air handling unit, a fan coil, and other equipment and connecting lines. The air conditioning system has a plurality of related control parameters, such as the starting and stopping of a water chilling unit and matched equipment, the outlet water temperature of chilled water, the inlet water temperature of cooling water, the frequency of a water pump and a fan and the like. If the control parameters are set by operation and maintenance personnel according to experience, unreasonable operation of the air conditioning system is probably caused, the problem that the building cannot meet the expected indoor environment (for example, the indoor temperature of the building cannot meet the expected temperature) and the problem of energy waste of the air conditioning system occur, and the energy saving and emission reduction effects are influenced.

Based on the above situation, with the development of cloud computing, the cloud server at the cloud end can be utilized in the embodiment of the present application, and based on at least real-time environment information of the building and future load information of the equipment system (real-time operation state information of the equipment system and the like can be further combined), when the building meets an expected indoor environment and the equipment system meets the energy consumption requirement, a control strategy of the equipment system is optimized; and the equipment system is controlled through the optimized control strategy, so that the energy-saving and emission-reducing effects of the building are improved.

It should be noted that one control strategy of the plant system may include: a plurality of control parameters of the plant system and control values for the respective control parameters. The control values of the control parameters in different control strategies may be different, so that a plurality of control parameters of different control values may be combined to form different control strategies. For convenience of understanding, taking an air conditioning system as an example, assuming that control parameters of the air conditioning system can be divided into three parts, namely start and stop of a refrigerator and supporting equipment, outlet water temperature of chilled water, inlet water temperature of cooling water and frequency of a water pump and a fan, the three parts of control parameters can be combined to obtain a control strategy, and the combined control strategies are different as long as control values of any control parameter in the three parts of control parameters are different.

Fig. 1 shows an alternative flowchart of a control optimization method for a device according to an embodiment of the present application. Optionally, the method flow may be implemented by the cloud server. Referring to fig. 1, the method flow may include the following steps.

In step S110, real-time attribute information is determined based on at least real-time environmental information of the building.

In some embodiments, the environmental information of the building may include external environmental information and indoor environmental information of the building. The external environment information of the building may be considered weather information (e.g., weather information, weather forecast information) and the like of the site where the building is located. The indoor environment information of the building may be, for example, temperature, humidity, carbon dioxide concentration, etc. in the building. The embodiment of the present application may determine the attribute information as the basic data based on at least the environmental information of the building (e.g., the external environmental information and the indoor environmental information of the building). When the control of the equipment system is optimized, the real-time attribute information needs to be utilized, so that the real-time attribute information can be determined at least based on the real-time environment information of the building. As an alternative implementation, the real-time environment information of the building may include real-time external environment information and indoor environment information of the building.

In further embodiments, the attribute information may also incorporate attributes of the building, as well as operational status information of the equipment system. The attributes of the building refer to factors of the building other than the environment (external environment and indoor environment), such as the building characteristics of the building, the number of people, the electric load, the production plan, and the like. The operation state information of the equipment system may be operation state information of the equipment system during operation, such as start-stop parameters of the equipment system, operation state parameters of various devices in the equipment system, system setting parameters of the equipment system, and the like. As an optional implementation, the embodiment of the application can acquire real-time environment information of a building, current attributes of the building and real-time running state information of an equipment system to determine the real-time attribute information; for example, the real-time attribute information may include: real-time environmental information of the building, current attributes of the building, and real-time operational status information of the equipment system.

As an optional implementation, the cloud server may obtain real-time external environment information of the building through a weather website, and may obtain real-time indoor environment information of the building, building characteristics, the number of people, power consumption loads, production plans, and the like through a building equipment automatic control system, a corresponding database, and the like.

As an optional implementation, a sensor for acquiring the running state information can be arranged in the equipment system, and the running state information acquired by the sensor can be transmitted to the cloud server; and the cloud server can acquire the real-time running state information of the equipment system based on the running state information acquired and transmitted by the sensor in real time. By taking an air conditioning system as an example, in the embodiment of the application, real-time running state information such as real-time freezing water flow, real-time freezing water inlet and return temperature, real-time cooling water inlet and return temperature, real-time freezing host set water outlet temperature and the like of the air conditioning system can be acquired in real time through the sensor.

In step S111, load prediction information of an equipment system for adjusting an indoor environment of the building at a future time is predicted according to the real-time attribute information.

The cloud server, after acquiring real-time attribute information (e.g., real-time environmental information of a building, real-time operation state information of an equipment system, etc.) as basic data, may predict load information (referred to as load prediction information) of the equipment system at a future time using the real-time attribute information. For example, load forecast information of the forecast equipment system in a future period of hours (such as load forecast information of the forecast equipment system in 3-5 hours in the future) may be set according to the situation, and the embodiment of the present application is not limited.

In some embodiments, the load prediction model in the form of a neural network can be constructed, and the training load prediction model is learned based on a machine learning method, so that the trained load prediction model has the following capabilities: load information of the facility system for a certain period of time after a certain time is predicted based on attribute information at the certain time (for example, environment information of a building at the certain time, operation state information of the facility system at the certain time, and the like). For example, the embodiment of the application may acquire historical attribute information at the historical time (e.g., environmental information of a building at the historical time, operating state information of an equipment system at the historical time, etc.), and historical load information at a period of time (e.g., within 3-5 hours) after the historical time; taking historical attribute information of the historical moment as training data, taking historical load information of a period of time after the historical moment, which is predicted by the load prediction model, as a training target, and training the load prediction model; the trained load prediction model may then have the above capabilities. As an optional implementation, the cloud server may input the real-time attribute information of the current time based on the trained load prediction model, so that the load prediction model predicts and obtains load prediction information of the equipment system in a future one-hour period after the current time.

In other embodiments, the load information of the equipment system at the current moment can be determined according to the real-time attribute information at the current moment; and predicting the load prediction information of the equipment system at the future time by combining the historical load trend of the equipment system and the load information at the current time. For example, the load information of the equipment system can be predicted periodically, and when the load prediction information of the equipment system in the next period in the future needs to be predicted, the load prediction information of the equipment system in the next period can be predicted according to the actual load information of the equipment system in the last period which has passed and the load information of the equipment system at the current moment. The time duration of the period (e.g. half an hour, 1 hour, etc.) may be set according to practical situations, and the embodiment of the present application is not limited.

It should be noted that there are many ways to predict the load prediction information of the equipment system at the future time, and the above-mentioned means are only optional means of the embodiment of the present application; in possible implementation, the load prediction information of the equipment system at the future time can be predicted by using computer simulation calculation, a regression analysis prediction method, a time series prediction method, an artificial neural network prediction method and the like based on the real-time attribute information, and a specific prediction means of the load prediction information is not limited in the embodiment of the application.

In step S112, determining a target control strategy by using an optimized control strategy model based on at least the real-time attribute information and the load prediction information, where the target control strategy is a control strategy in which a building indoor environment meets an expected indoor environment and an equipment system energy consumption meets an energy consumption requirement among a plurality of control strategies; the optimization control strategy model is obtained based on simulation model training, and the simulation model is based on attribute information and a control strategy, and the building indoor environment and the energy consumption of the equipment system are obtained through simulation calculation.

In some embodiments, the optimization control strategy model can be obtained through learning and training in advance, and the optimization control strategy can determine the building indoor environment and the energy consumption of the equipment system respectively corresponding to the multiple control strategies at least based on the real-time attribute information and the load prediction information, so that the target control strategy that the building indoor environment meets the expected indoor environment and the energy consumption of the equipment system meets the energy consumption requirement is optimized from the multiple control strategies. The optimal control strategy model provided by the embodiment of the application can be obtained by learning and training based on a simulation model, for example, the optimal control strategy model is obtained by reinforcement learning and training based on the simulation model; for another example, training data for training the optimal control strategy model is obtained based on simulation model simulation, so that the optimal control strategy model is obtained by training with the training data in a machine learning manner.

In the embodiment of the application, the simulation model may simulate an operation process of the equipment system, and under the condition that at least a control strategy is provided, the simulation model may simulate different attribute information (for example, environment information different from a building, operation state information different from the equipment system, and the like), and obtain the energy consumption of the equipment system and the indoor environment of the building through simulation calculation. Taking the air conditioning system as an example, the simulation model can simulate the operation process of the air conditioning system, when the control strategy is provided, the simulation model can simulate different attribute information, and the energy consumption of the air conditioning system and the indoor environment temperature of the building are obtained through simulation calculation.

It should be noted that although the simulation model can obtain the building indoor environment and the equipment system energy consumption corresponding to the control strategy through simulation calculation, the simulation model is subjected to simulation calculation based on the operation of the analog equipment system, which results in a large calculation amount of the simulation model, and particularly when the control parameters of the equipment system are large, the number of the control strategies obtained by combining the control parameters of different control values is large, which aggravates the complexity of the simulation model for simulating the operation of the equipment system under different control strategies, and leads to a rapid increase of the calculation amount of the simulation calculation. Therefore, if the simulation model is directly deployed on line to determine the building indoor environments and the equipment system energy consumption corresponding to different control strategies, the calculation amount of data on line is very large, and the calculation efficiency on line is reduced. Based on the above, in the embodiment of the application, the simulation model is not deployed on line, but the optimization control strategy model is learned and trained through the simulation model, and then the optimization control strategy model is deployed on line, and the optimization control strategy model determines the building indoor environment and the equipment system energy consumption corresponding to each control strategy at least based on the real-time attribute information and the load prediction information. In the embodiment of the application, the optimized control strategy model does not obtain the building indoor environment and the equipment system energy consumption corresponding to each control strategy through simulation calculation, but predicts and outputs the building indoor environment and the equipment system energy consumption corresponding to each control strategy based on the pre-learning and training effects. The optimization control strategy model can also be called a proxy model of the simulation model, namely the optimization control strategy model proxy simulation model is deployed on the line.

In some embodiments, the training data for training the optimal control strategy model can be obtained based on simulation of the simulation model, for example, the simulation model can simulate and calculate indoor environment of a building and energy consumption of an equipment system under different attribute information and different control strategies; therefore, training of the optimal control strategy model is carried out by taking different attribute information and different control strategies as training data and taking the building indoor environment and equipment system energy consumption obtained through simulation as training labels. As an optional training implementation manner, in the embodiment of the application, the attribute information and the control strategy may be input into the optimal control strategy model, and the optimal control strategy model is trained by using the building indoor environment and the equipment system energy consumption predicted and output by the optimal control strategy model, as a training target, the building indoor environment and the equipment system energy consumption approaching to the corresponding simulation calculation of the simulation model. After further training, the optimization control strategy model can predict and obtain the building indoor environment and the energy consumption of the equipment system based on the input attribute information, the load prediction information and the control strategy.

In the process of training the optimal control strategy model in a machine learning mode, the training data comes from the simulation of the simulation model, so that the range of the training data can be enlarged, the condition that the real data is not enough for training is avoided, and the training limit of the optimal control strategy model is reduced. In some embodiments, the optimization control strategy model may be a structure of a neural network, and the machine learning manner may be, for example, algorithms such as logistic regression, GBDT (Gradient Boosting Decision Tree), support vector machine, and neural network.

In the case of training the optimal control strategy model in the machine learning manner, as an optional implementation of step S112, in the embodiment of the present application, at least the real-time attribute information, the load prediction information, and the plurality of control strategies may be input into the optimal control strategy model, so as to obtain the building indoor environment and the equipment system energy consumption respectively corresponding to the plurality of control strategies predicted by the optimal control strategy model. That is, the optimized control strategy model may predict the building indoor environment and the equipment system energy consumption corresponding to the output under each control strategy. Furthermore, the method and the device for controlling the indoor environment of the building can determine the target control strategy that the indoor environment of the building meets the expected indoor environment and the energy consumption meets the energy consumption requirement based on the indoor environment of the building and the energy consumption of the equipment system respectively corresponding to the plurality of control strategies. In some embodiments, the energy consumption requirements of the expected indoor environment of the building and the equipment system may be set, so that the indoor environment of the building is optimized to meet the expected indoor environment in the plurality of control strategies, the energy consumption of the equipment system meets the energy consumption requirement of the control strategy, and the optimized control strategy is used as the target control strategy. The target control strategy may be regarded as a control strategy for controlling the plant system that is finally executed. As an alternative implementation, the embodiment of the present application may define a control strategy with the lowest energy consumption and a building indoor environment that meets the expected indoor environment as a target control strategy. Taking the air conditioning system as an example, the embodiment of the application can determine that the indoor temperature of the building reaches the expected temperature, and the control strategy with the lowest energy consumption of the air conditioning system is the target control strategy.

In some embodiments, the simulation model may also be used as a simulator to train the optimal control strategy model through a reinforcement learning algorithm. Reinforcement learning, also known as refinish learning, evaluative learning, or reinforcement learning, is used to describe and solve the problem of an agent in interacting with the environment to achieve maximum return or achieve a specific goal through a learning strategy. Furthermore, after reinforcement learning, the optimal control strategy model may predict, based on the attribute information and the load prediction information, that the building indoor environment meets an expected indoor environment from among the plurality of control strategies and that the energy consumption of the equipment system meets the target control strategy of the energy consumption requirement. In the case of learning and training the optimal control strategy model in a reinforcement learning manner, as an optional implementation of step S112, in the embodiment of the present application, the real-time attribute information and the load prediction information may be input into the optimal control strategy model, and the building indoor environment predicted by the optimal control strategy model from the multiple control strategies is obtained to meet the expected indoor environment, and the energy consumption of the equipment system meets the target control strategy of the energy consumption requirement.

In some embodiments, optimizing the plurality of control strategies used by the control strategy model for optimization may be based on a grid search method. As an alternative implementation, the plurality of control strategies are obtained by searching the control values of the plurality of control parameters using a grid search method, and the plurality of control parameters of different control values are combined to obtain different control strategies. For example, in the embodiment of the present application, control parameters of discrete values such as switches and gears of an equipment system may be exhausted, control parameters of continuous values such as a temperature setting value may be sampled at equal intervals (or other sampling strategies at unequal intervals may be adopted), and then all the control parameters are arranged and combined, so as to obtain a plurality of control strategies searched by using a grid search method. In other possible embodiments, the embodiments of the present application may also adopt a heuristic search method to obtain a plurality of control strategies input into the optimized control strategy model.

In other embodiments, the plurality of control strategies input to the optimal control strategy model may be a plurality of random control strategies. For example, by determining random control values of the respective control parameters, a plurality of control parameters having random control values are combined, so that a plurality of control parameters of different random control values can be combined to obtain different random control strategies.

In some further embodiments, based on the determined target control policy, the cloud server may transmit control information carrying the target control policy to the device system or a control system of the device system, so that the device system sets various control parameters of the device system based on the target control policy, and the device system is enabled to work under the control of the target control policy, so that the building is close to and meets the expected indoor environment, and the device system is close to and meets the energy consumption requirement. In some possible embodiments, the cloud server may also issue the target control policy to the operation and maintenance staff, and the operation and maintenance staff sets the control parameters of the device system based on the target control policy.

It should be noted that, in the embodiment of the present application, optimization of the control strategy can be performed according to the building indoor environment and the energy consumption of the equipment system in a plurality of control strategies, and one control strategy is obtained by combining a plurality of control parameters, so that in the embodiment of the present application, an optimized target control strategy can be provided under the condition that a plurality of control parameters of the whole equipment system are considered in a unified manner, thereby achieving global optimization of the control strategy in a plurality of control strategies of control parameter combinations with different control values, achieving an optimization effect that the optimized target control strategy can achieve global control parameters, and avoiding trapping in local optimization of the control parameters.

In some embodiments, the target control strategy may be a control strategy for the plant system at a future time. That is to say, the embodiment of the application not only considers the current real-time attribute information, but also predicts the load of the equipment system at the future time, so that the target control strategy of the equipment system at the future time can be given based on the load prediction result at the future time.

According to the control optimization method for the equipment, the simulation model is not deployed on line (the simulation model carries out simulation calculation based on the operation process of a simulation equipment system), but the optimization control strategy model obtained based on the simulation model training is deployed on line, and the optimization control strategy model determines the building indoor environment and the equipment system energy consumption corresponding to different control strategies in a machine prediction mode, so that huge calculation amount caused by the fact that the simulation model is directly used on line can be avoided, and the efficiency of on-line calculation is improved. Based on the real-time attribute information, the cloud server can determine the real-time attribute information at least based on the real-time environment information of the building, and predict the load prediction information of the equipment system at the future time according to the real-time attribute information; therefore, the cloud server can determine a target control strategy by using an optimization control strategy model at least based on the real-time attribute information and the load prediction information, wherein the target control strategy is a control strategy in which the building indoor environment meets the expected indoor environment and the energy consumption of the equipment system meets the energy consumption requirement in a plurality of control strategies. According to the embodiment of the application, the equipment system can be controlled under the condition that the building meets the expected indoor environment, so that the equipment system meets the energy consumption requirement, and the energy-saving and emission-reducing effects of the building are improved.

To facilitate understanding of the solutions provided by the embodiments of the present application, fig. 2 exemplarily illustrates an alternative example diagram of a system architecture of the embodiments of the present application. As shown in fig. 2, the system architecture may include: the system comprises a building 210, a device system 220 for regulating the building 210, a plurality of sensors 230 for collecting the running state information of the device system 220, a weather website 240 for providing external weather information of the building 210, a building equipment automatic control system 250 of the building 210, and a cloud server 260 arranged in the cloud.

The plurality of sensors 230 may collect the operation state information of the device system 220 in real time, and transmit the real-time operation state information to the cloud server 260. The cloud server 260 may acquire real-time external environmental information (e.g., weather information, weather forecast information, etc. of a region where the building is located), real-time indoor environmental information, and current attributes (e.g., building characteristics, number of people, power load, production plan, etc.) of the building 210 from the weather website 240, the building automation system 250, a database of the building, and the like. For convenience of explanation, real-time external environment information, real-time indoor environment information, current attributes of the building 210, and operation state information of the equipment system 220 may be used as the real-time attribute information. The cloud server 260 may store the acquired real-time attribute information in a database of the cloud server. It should be noted that the cloud server may also obtain part or all of the real-time attribute information from other devices or a database, which may be determined according to the circumstances, and the embodiments of the present application are not limited.

A load prediction model 261 and an optimization control strategy model 262 may be provided in the cloud server 260. The cloud server may store the real-time attribute information in a database and input the load prediction model 261, thereby obtaining load prediction information of the equipment system at a future time output by the load prediction model 261. Further, the cloud server 260 may input at least the real-time attribute information and the load prediction information into the optimal control strategy model 262, so as to obtain the building indoor environment and the equipment system energy consumption corresponding to the plurality of control strategies, respectively. The cloud server 260 may optimize a target control strategy in which the building indoor environment meets an expected indoor environment and the energy consumption of the equipment system meets the energy consumption requirement among a plurality of control strategies.

The cloud server 260 may transmit the control information carrying the target control policy to the equipment system 220, or the building equipment automation system 250, or the operation and maintenance personnel of the equipment system 220. Furthermore, the equipment system 220 is set with control parameters according to a target control strategy, so that the work of the equipment system 220 can meet the energy consumption requirement, the building can reach the effect of expected indoor environment, and the effects of energy conservation and emission reduction are improved.

As an alternative implementation, fig. 3 exemplarily illustrates an exemplary diagram of a model training phase provided by an embodiment of the present application. As shown in fig. 3, the model training phase may include: a load prediction model training stage 310, a simulation model establishing stage 320, an optimization control strategy model establishing stage 330 and a model on-line deployment stage 340.

In the stage 310 of training the load prediction model, the embodiment of the present application may train to obtain the load prediction model based on the historical attribute information at the historical time. For example, the embodiment of the application may store historical attribute information at a historical time (e.g., historical environmental information of a building at the historical time, historical operating state information of an equipment system at the historical time), and load information of the equipment system for a period of time after the historical time by the database; therefore, the embodiment of the application can acquire historical attribute information of the historical moment and load information of the equipment system after the historical moment from the database; furthermore, in the embodiment of the application, historical attribute information of the equipment system at the historical time is used as training data, load information of the equipment system after the historical time is used as a training label, and the load prediction model is trained, so that the trained load prediction model can predict the load prediction information of the equipment system after the current time in a future time based on the real-time attribute information of the current time.

In the simulation model establishing stage 320, a simulation model may be established in the embodiment of the present application, so that the simulation model can simulate the operation process of the equipment system, simulate different attribute information under a control strategy under the condition of providing at least the control strategy, and obtain the energy consumption of the equipment system and the indoor environment of the building through simulation calculation. In some embodiments, the simulation model may be created by using simulation software, such as energy plus (a building energy consumption simulation engine), Dest (a software platform for simulation of building environment and HVAC System, HVAC refers to heating ventilation and air conditioning), modecia (an open, object-oriented, equation-based computer language, which may span different fields and conveniently implement modeling of complex physical systems), trsys (Transient System simulation program), and the like.

As an optional implementation of establishing the simulation model by using the simulation software, the embodiment of the application may establish the system energy consumption simulation model of the equipment system by using the simulation software based on the attributes of the building, the device combination condition of the equipment system, the equipment information of the equipment system, and the like. In some embodiments, the simulation model of the plant system may be formed from an energy consumption model of a plurality of parameters of the plant system, one parameter may be considered to be an important device or combination of devices in the plant system; taking an air conditioning system as an example, the simulation model of the air conditioning system can be formed by energy consumption models of a plurality of parameters such as a water chilling unit energy consumption model, a cooling tower and fan energy consumption model, a chilled water temperature control model and the like. As an optional implementation, the energy consumption models of all parameters can be modularized as an integral simulation model, and the energy consumption models of different parameters can be connected through a data interface. In further embodiments, after the simulation model is established, the system parameters of the plant system may be identified and the simulation model may be calibrated based on actual operation data of the plant system during actual operation.

In the stage 330 of building the optimal control strategy model, the embodiment of the present application may obtain the optimal control strategy model based on the simulation model training. As an optional implementation, the optimization control strategy model can be obtained by learning and training an intelligent algorithm by using a simulation model. For example, the embodiment of the application can utilize a simulation model to generate a large amount of training data, and based on the data, algorithms such as logistic regression, GBDT, support vector machine, neural network and the like are adopted to train machine learning models in forms such as neural network and the like, so as to obtain an optimal control strategy model. For another example, in the embodiment of the application, the simulation model can be used as a simulator, and the reinforcement learning model is trained through a reinforcement learning algorithm to obtain the optimization control strategy model.

Taking an optimized control strategy model obtained by machine learning training as an example, fig. 4 shows a flowchart of an optional method for training the optimized control strategy model provided in the embodiment of the present application. As shown in fig. 4, the method flow may include:

in step S410, the building indoor environment and the energy consumption of the equipment system under different attribute information and different control strategies are calculated through simulation of the simulation model.

In step S411, the different attribute information and the different control strategies are used as training data, and the building indoor environment and the equipment system energy consumption predicted and output by the optimal control strategy model approach to the building indoor environment and the equipment system energy consumption calculated by the simulation model are used as training targets, so as to train the optimal control strategy model.

The simulation model can simulate different attribute information based on the control strategy, and simulate and calculate to obtain the building indoor environment and the equipment system energy consumption, so that the simulation obtains the building indoor environment and the equipment system energy consumption under different attribute information and different control strategies. Alternatively, the control strategy provided to the simulation model may be a stochastic control strategy.

In some embodiments, all possible attribute information (for example, annual building environment information, running state information of an equipment system, and the like) can be pre-recorded in a simulation model, so that under the condition of providing a control strategy for the simulation model, the simulation model can simulate different attribute information based on pre-recorded data, and simulate and calculate to obtain building indoor environments and equipment system energy consumption corresponding to the different attribute information; furthermore, by providing different control strategies for the simulation model, the building indoor environment and the equipment system energy consumption under different attribute information and different control strategies can be obtained. In other embodiments, the attribute information and the control strategy can also be used as input data and input to the simulation model, so that the simulation model can simulate the building indoor environment and the energy consumption of the equipment system according to the input attribute information under the input control strategy. Furthermore, by inputting different attribute information and control strategies into the simulation model, the building indoor environment and equipment system energy consumption under different attribute information and different control strategies can be obtained.

After the simulation model performs simulation calculation to obtain different attribute information, building indoor environments under different control strategies and equipment system energy consumption, the embodiment of the application can take the different attribute information and the different control strategies as training data of the optimization control strategy model to train the optimization control strategy model; in the training process, the parameters of the optimization control strategy model are adjusted by taking the indoor environment of the building and the energy consumption of the equipment system which are predicted and output by the optimization control strategy model and approach to the indoor environment of the building and the energy consumption of the equipment system obtained through simulation calculation as a training target, so that the indoor environment of the building and the energy consumption of the equipment system can be predicted and obtained by the trained optimization control strategy model based on the attribute information, the load prediction information and the control strategy.

In further embodiments, when the optimization control strategy model is trained, load prediction information may also be used as a training input in the embodiments of the present application, and the load prediction information may be obtained from an established load prediction model, or may be calculated by using a simulation model to obtain future load prediction information.

As an optional implementation, in the training process of the optimization control strategy model, the input of the optimization control strategy model may be attribute information (building environment information, equipment system operating state information, and the like), a control strategy, and load prediction information, and the output may be the predicted building indoor environment and equipment system energy consumption under each control strategy.

After the optimization control strategy model is trained, the input of the optimization control strategy model can be real-time attribute information, control strategies and load prediction information, and the prediction output of the optimization control strategy model can be the predicted indoor environment of the building and the energy consumption of the equipment system under each control strategy. Taking an air conditioning system as an example, the input of the optimization control strategy model can be the environmental information of a building, the future load, the running state information and the control strategy of the air conditioning system, and the output can be the energy consumption of the air conditioning system and the indoor temperature of the building.

The training data of the training optimization control strategy model in the embodiment of the application is derived from simulation of the simulation model instead of being based on historical real data, so that the condition that the real data samples are less and are not used for training is avoided, and the problem of poor generalization of the model caused by limitation of the training data is avoided. Taking an air conditioning system as an example, the air conditioning system is complex in structure and numerous in influencing factors, a relatively accurate optimal control strategy model needs to be obtained through training, a large amount of diverse operation data of the air conditioning system is needed, and in the actual operation process of the air conditioning system, operation and maintenance personnel often have a fixed mode and low operation frequency, so that fewer effective samples are obtained in the actual operation of the air conditioning system, and therefore the optimal control strategy model is trained based on the actual operation data of the air conditioning system, and the generalization of the trained optimal control strategy model is poor. Further, because the characteristic differences of different air conditioning systems are large, even if a large amount of operation data of different air conditioning systems are collected to train the optimization control strategy model, the generalization of the model is also limited. Based on this, the embodiment of the application generates the operation data of the air conditioning system under a large number of random control strategies through the simulation model, thereby avoiding the problem of poor model generalization caused by the limitation of training data when the optimal control strategy model is trained, and promoting the generalization of the trained optimal control strategy model.

In some embodiments, the simulation model should meet at least the following requirements: supporting dynamic simulation; the real-time output of data such as load, energy consumption, attribute information and the like in the simulation process is supported; and real-time input or adjustment of the control strategy in the simulation process is supported. In some further embodiments, because the simulation model is obtained by performing simulation calculation in a mode of simulating the operation of the equipment system, the energy consumption of the equipment system and the indoor environment of the building are obtained, so that the simulation calculation result of the simulation model may have a certain difference from the real situation, and in order to ensure the accuracy of the optimal control strategy model obtained by training based on the simulation model, the embodiment of the application can calibrate the simulation model by using the actual operation data of the equipment system during the operation in the stage of establishing the simulation model, thereby improving the accuracy of the simulation model and ensuring the accuracy of the optimal control strategy model trained based on the simulation model.

In some embodiments, after the optimal control strategy model is deployed online, the embodiment of the application can further perform update training on the optimal control strategy model by using real data, so that the accuracy of the optimal control strategy model is improved. For example, after the optimal control strategy model is on line and the device system executes control optimization based on the target control strategy, the embodiment of the present application may obtain the actual device system energy consumption and the actual building indoor environment under the target control strategy, and update and train the optimal control strategy model according to the actual building indoor environment and the actual device system energy consumption, so as to update the parameters of the optimal control strategy model, so that the building indoor environment and the device system energy consumption output by the optimal control strategy model are close to the actual situation (for example, the actual building indoor environment and the actual device system energy consumption).

In some further embodiments, the optimal control strategy model is obtained by training in a machine learning manner or a reinforcement learning manner, and may be based on empirical analysis, data analysis, or rules obtained by simulation model experiments. In a possible implementation, the optimization control strategy model may be a rule-based algorithm formula, control logic code, etc., which may express an optimization relationship between attribute information (e.g., environmental information of a building, operational state information of an equipment system, etc.), load prediction information, a control strategy, an indoor environment of a building, and energy consumption of the equipment system. For example, the optimization control strategy model may express the optimization relationship between the attribute information, the load prediction information, the control strategy, the building indoor environment and the energy consumption of the equipment system in the form of an algorithm formula, a control logic code, a rule description and the like on the basis of empirical analysis, data analysis or simulation model experiments. Furthermore, based on the optimization control strategy model in a regular form provided by the embodiment of the application, the embodiment of the application can obtain the target control strategy that the building indoor environment recommended by the optimization control strategy model conforms to the expected indoor environment and the energy consumption of the equipment system conforms to the energy consumption requirement by processing the algorithm formula, the control logic code and the like corresponding to the optimization control strategy model under the condition of inputting the attribute information, the load prediction information and the like.

In the on-line model deployment stage 340, the load prediction model and the optimal control strategy model are deployed on the cloud server on line in the embodiment of the present application.

In some embodiments, after the model is deployed online, during the actual operation of the equipment system, the online deployed model may be called at intervals (e.g., half an hour, etc.) to perform optimal control on the equipment system. For example, at intervals, the load prediction information is predicted by using the load prediction model, and the building indoor environment and the energy consumption corresponding to the equipment system under a plurality of control strategies are determined by using the optimal control strategy model based on the real-time attribute information and the corresponding load prediction information, so that a target control strategy that the building indoor environment meets the building indoor environment expectation and the energy consumption of the equipment system meets the energy consumption requirement is optimized, and the equipment system is controlled to work based on the target control strategy.

In some embodiments, after the cloud server determines the target control policy, the target control policy may be issued to the operation and maintenance personnel, and the operation and maintenance personnel manually set the control parameters of the equipment system based on the target control policy. In other embodiments, the cloud server may send the control information carrying the target control policy to the device system or a control system connected to the device system, so as to implement automatic control of the device system.

The control optimization scheme of the equipment system provided by the embodiment of the application not only considers the current real-time attribute information, but also predicts the load of the equipment system at the future time, and provides a target control strategy of the equipment system at the future time based on the load prediction result. That is, the target control strategy may be a control strategy of the plant system at a future time corresponding to the predicted load prediction information. Meanwhile, the embodiment adopts a global optimization method, takes the whole control parameters into consideration uniformly, gives an optimized control strategy combined by the control parameters, and avoids falling into the optimization of local control parameters. Meanwhile, the embodiment of the application generates a large amount of training data by introducing the simulation model, reduces the dependence on actual operation data, enables the optimization control strategy model to be sufficiently trained, and reduces the training limit of the optimization control strategy model. According to the embodiment of the application, the energy consumption of the equipment system and the indoor environment of the building under the control strategy are determined by performing online calculation through the optimization control strategy model, the calculation of the energy consumption of the equipment system and the indoor environment of the building in an online environment in a simulation mode is avoided, and the calculation efficiency is improved. According to the embodiment of the application, global optimization control of equipment systems (such as air conditioning systems) can be realized through high calculation efficiency, so that the building can meet the expected indoor environment, the equipment systems can meet the energy consumption requirement, and the energy-saving and emission-reducing effects are improved.

In further embodiments, after determining the target control policy of the device system, the cloud server may feed back the target control policy (including target values of a plurality of control parameters of the device system) to the presentation platform, so as to present the current values and the target values of the respective control parameters of the device system on the presentation platform, thereby facilitating a user (e.g., an operation and maintenance person) to view differences between the current values and the target values of the respective control parameters of the device system. Optionally, the display platform may be a monitoring large screen of a cloud server, a display interface of a terminal device, and the like, and the embodiment of the present application is not limited. As an alternative implementation, fig. 5A shows another alternative flowchart of a control optimization method of a device provided in an embodiment of the present application. The method flow may be implemented by the presentation platform, and as shown in fig. 5A, the method flow may include the following steps.

In step S510, a target control policy of the equipment system fed back by the cloud server is obtained, where the target control policy includes target values of a plurality of control parameters of the equipment system.

After determining the target control strategy of the equipment system, the cloud server can feed back the target control strategy to the display platform. In some embodiments, the cloud server may periodically update the target control policy of the equipment system, so as to periodically feed back the target control policy to the presentation platform, so that the presentation platform periodically updates the target values of the plurality of control parameters of the presented equipment system.

In an embodiment of the application, the target control strategy is determined by the cloud server by using an optimized control strategy based on at least real-time attribute information and load prediction information of the equipment system at a future time; the optimization control strategy model is obtained based on simulation model training, and the simulation model is based on attribute information and a control strategy, and the building indoor environment and the energy consumption of the equipment system are obtained through simulation calculation; the real-time attribute information is determined at least based on real-time environment information of a building, the load prediction information is obtained by prediction according to the real-time attribute information, and the equipment system is used for adjusting the indoor environment of the building. The detailed related contents described in this paragraph can refer to the descriptions of the corresponding parts previously, and are not described herein again.

In step S511, the current values and the target values of the plurality of control parameters are displayed on the optimization control interface of the equipment system, so that the user can view the difference between the current values and the target values of the respective control parameters of the equipment system.

The display platform can provide an optimized control interface of the equipment system, and based on the target control strategy of the equipment system obtained from the cloud server, the display platform can display the target values of the control parameters carried by the target control strategy on the optimized control interface. Meanwhile, the optimization control interface can also display the current values of a plurality of control parameters of the equipment system, so that a user can check the difference between the current values and the target values of the control parameters of the equipment system.

It can be seen that, in the embodiment of the present application, a display platform displays an optimized control interface of an equipment system, where the optimized control interface displays a target control strategy of the equipment system, where the target control strategy includes target values of a plurality of control parameters of the equipment system; the optimization control interface displays the current values of the control parameters;

In some embodiments, FIG. 5B illustrates a schematic diagram of an optimization control interface. As shown in FIG. 5B, the optimization control interface may present a plurality of control parameters (e.g., control parameters 1 through n), current values of the respective control parameters, and target values of the plant system. In further embodiments, the modes of adjustment of the control parameters of the plant system may be divided into automatic adjustment and manual adjustment; aiming at the automatically adjusted control parameters, the equipment system can automatically adjust the numerical values of the control parameters based on the target values; for the manually adjusted control parameters, a user (e.g., an operation and maintenance person) needs to decide whether to adjust the current values of the control parameters to the target values.

In further some embodiments, the cloud server may also transmit the indoor environment information of the building to the display platform in real time, so as to display the indoor environment information of the building on the display platform in real time, for example, the optimal control interface shown in fig. 5B may also display the indoor environment information of the building, such as the indoor temperature, the indoor CO content, and the like. In further embodiments, the cloud server may further determine information such as power consumption of the device system and send the information to the display platform, so that the display platform may further display the information such as power consumption of the device system on the optimized control interface.

In the following, a control optimization apparatus of an equipment system provided in the embodiment of the present application is introduced, and the apparatus content described below may be regarded as a functional module that is required by a cloud server to implement the control optimization method of the equipment provided in the embodiment of the present application. The device content described below may be referred to in correspondence with the method content described above.

Fig. 6 exemplarily shows an alternative block diagram of a control optimization device of an equipment system provided by an embodiment of the present application. Referring to fig. 6, the apparatus may include:

an information determination module 610 for determining real-time attribute information based at least on real-time environmental information of the building;

a load prediction module 611, configured to predict load prediction information of an equipment system at a future time according to the real-time attribute information, where the equipment system is configured to adjust an indoor environment of the building;

an optimizing module 612, configured to determine a target control strategy by using an optimized control strategy model based on at least the real-time attribute information and the load prediction information, where the target control strategy is a control strategy in which a building indoor environment meets an expected indoor environment and an equipment system energy consumption meets an energy consumption requirement among a plurality of control strategies; the optimization control strategy model is obtained based on simulation model training, and the simulation model is based on attribute information and a control strategy, and the building indoor environment and the energy consumption of the equipment system are obtained through simulation calculation.

In some embodiments, the optimizing module 612, configured to determine the target control strategy using the optimization control strategy model based on at least the real-time attribute information and the load prediction information, includes:

inputting at least the real-time attribute information, load prediction information, and a plurality of control strategies into an optimal control strategy model; acquiring building indoor environments and equipment system energy consumption respectively corresponding to a plurality of control strategies predicted by an optimized control strategy model; and determining a target control strategy that the building indoor environment meets the expected indoor environment and the energy consumption meets the energy consumption requirement based on the building indoor environment and the equipment system energy consumption respectively corresponding to the plurality of control strategies.

In some further embodiments, the apparatus provided in this application may further be configured to:

building indoor environments and equipment system energy consumption under different attribute information and different control strategies are calculated through simulation of a simulation model; and taking different attribute information and different control strategies as training data, taking the building indoor environment and equipment system energy consumption predicted and output by the optimization control strategy model as training targets, and training the optimization control strategy model, wherein the building indoor environment and the equipment system energy consumption are close to the building indoor environment and the equipment system energy consumption calculated by the simulation model.

In some embodiments, the apparatus provided in this application, for the simulation calculation through the simulation model, the building indoor environment and the equipment system energy consumption under different attribute information and different control strategies includes:

simulating the operation process of the equipment system based on different attribute information and different control strategies through a simulation model; and acquiring the operation process of the simulation model based on the simulated equipment system, the building indoor environment of the simulation calculation and the energy consumption of the equipment system.

In some embodiments, the simulation model is built using simulation software; after the simulation model is established, the system parameters of the equipment system are identified and the simulation model is calibrated according to the actual operation data of the equipment system.

inputting the real-time attribute information and the load prediction information into an optimization control strategy model, and acquiring a target control strategy, wherein the building indoor environment predicted by the optimization control strategy model from a plurality of control strategies meets the expected indoor environment, and the energy consumption of the equipment system meets the energy consumption requirement; the optimization control strategy model is obtained by training the simulation model through a reinforcement learning algorithm.

In some embodiments, the plurality of control strategies are obtained by searching the control values of the plurality of control parameters by using a grid search method, and the plurality of control parameters with different control values are combined to obtain different control strategies; or the plurality of control strategies comprise a plurality of random control strategies, and a plurality of control parameters with different random control values are combined to obtain different random control strategies.

In some embodiments, the load prediction module 611 is configured to predict load prediction information of the equipment system at a future time according to the real-time attribute information, and includes:

and inputting the real-time attribute information into a load prediction model, and acquiring load prediction information of the equipment system predicted by the load prediction model at the future time.

acquiring historical attribute information of a historical moment and load information of a device system after the historical moment; and training a load prediction model by taking the historical attribute information as training data and taking the load information of the equipment system after the historical moment as a training label.

In some embodiments, the information determining module 610, configured to determine the real-time attribute information based on at least the real-time environmental information of the building, includes: the method comprises the steps of obtaining real-time environment information of a building, current attributes of the building and real-time running state information of an equipment system to determine the real-time attribute information.

and sending the target control strategy to operation and maintenance personnel of the equipment system, or sending control information carrying the target control strategy to the equipment system or a control system of the equipment system, so that the equipment system sets control parameters based on the target control strategy and works.

and acquiring the actual indoor environment of the building and the energy consumption of the actual equipment system under the target control strategy, and updating and training the optimization control strategy model according to the actual indoor environment of the building and the energy consumption of the actual equipment system.

The embodiment of the application also provides a cloud server, and the cloud server can execute the control optimization method of the equipment provided by the embodiment of the application. Fig. 7 illustrates an alternative block diagram of a cloud server. As shown in fig. 7, the cloud server may include: at least one processor 71, at least one communication interface 72, at least one memory 73 and at least one communication bus 74.

In the embodiment of the present application, the number of the processor 71, the communication interface 72, the memory 73 and the communication bus 74 is at least one, and the processor 71, the communication interface 72 and the memory 73 are communicated with each other through the communication bus 74.

Alternatively, the communication interface 72 may be an interface of a communication module for performing network communication.

Alternatively, the processor 71 may be a CPU (central Processing Unit), a GPU (Graphics Processing Unit), an NPU (embedded neural network processor), an FPGA (Field Programmable Gate Array), a TPU (tensor Processing Unit), an AI chip, an asic (application Specific Integrated circuit), or one or more Integrated circuits configured to implement the embodiments of the present application.

The memory 73 may comprise high-speed RAM memory, and may also include non-volatile memory (non-volatile memory), such as at least one disk memory.

The memory 73 stores one or more computer-executable instructions, and the processor 71 calls the one or more computer-executable instructions to execute the control optimization method of the device provided by the embodiment of the present application.

Embodiments of the present application also provide a storage medium, which may store one or more computer-executable instructions that, when executed, implement a control optimization method of a device as provided in embodiments of the present application.

Embodiments of the present application further provide a computer program, which when executed, implements the method for controlling and optimizing the device provided in the embodiments of the present application.

While various embodiments have been described above in connection with what are presently considered to be the embodiments of the disclosure, the various alternatives described in the various embodiments can be readily combined and cross-referenced without conflict to extend the variety of possible embodiments that can be considered to be the disclosed and disclosed embodiments of the disclosure.

Although the embodiments of the present application are disclosed above, the present application is not limited thereto. Various changes and modifications may be effected therein by one of ordinary skill in the pertinent art without departing from the scope or spirit of the present disclosure, and it is intended that the scope of the present disclosure be defined by the appended claims.

Claims

1. A method for control optimization of a plant, comprising:

2. The plant control optimization method of claim 1, wherein the determining a target control strategy using an optimized control strategy model based on at least the real-time attribute information and the load forecast information comprises:

inputting at least the real-time attribute information, load prediction information, and a plurality of control strategies into an optimal control strategy model; acquiring building indoor environments and equipment system energy consumption respectively corresponding to a plurality of control strategies predicted by an optimized control strategy model;

and determining a target control strategy that the building indoor environment meets the expected indoor environment and the energy consumption meets the energy consumption requirement based on the building indoor environment and the equipment system energy consumption respectively corresponding to the plurality of control strategies.

3. The control optimization method of the plant according to claim 2, further comprising:

building indoor environments and equipment system energy consumption under different attribute information and different control strategies are calculated through simulation of a simulation model;

and taking different attribute information and different control strategies as training data, taking the building indoor environment and equipment system energy consumption predicted and output by the optimization control strategy model as training targets, and training the optimization control strategy model, wherein the building indoor environment and the equipment system energy consumption are close to the building indoor environment and the equipment system energy consumption calculated by the simulation model.

4. The control optimization method of the equipment according to claim 3, wherein the simulation calculation through the simulation model, the indoor environment of the building and the energy consumption of the equipment system under different attribute information and different control strategies comprises:

simulating the operation process of the equipment system based on different attribute information and different control strategies through a simulation model;

acquiring the operation process of a simulation model based on a simulated equipment system, the building indoor environment of simulation calculation and the energy consumption of the equipment system;

wherein the simulation model is established by using simulation software; after the simulation model is established, the system parameters of the equipment system are identified and the simulation model is calibrated according to the actual operation data of the equipment system.

5. The plant control optimization method of claim 1, wherein the determining a target control strategy using an optimized control strategy model based on at least the real-time attribute information and the load forecast information comprises:

6. The control optimization method for the plant according to any one of claims 1 to 5, wherein the plurality of control strategies are obtained by searching control values of a plurality of control parameters using a grid search method, and the plurality of control parameters of different control values are combined to obtain different control strategies; or the plurality of control strategies comprise a plurality of random control strategies, and a plurality of control parameters with different random control values are combined to obtain different random control strategies.

7. The plant control optimization method of claim 1, wherein said predicting load forecast information for a plant system at a future time based on said real-time attribute information comprises:

inputting the real-time attribute information into a load prediction model, and acquiring load prediction information of an equipment system predicted by the load prediction model at future time;

the method further comprises the following steps:

acquiring historical attribute information of a historical moment and load information of a device system after the historical moment;

and training a load prediction model by taking the historical attribute information as training data and taking the load information of the equipment system after the historical moment as a training label.

8. The control optimization method of equipment of claim 1, wherein the determining real-time attribute information based on at least real-time environmental information of a building comprises:

acquiring real-time environment information of a building, current attributes of the building and real-time running state information of an equipment system to determine the real-time attribute information;

the method further comprises the following steps:

sending the target control strategy to operation and maintenance personnel of an equipment system, or sending control information carrying the target control strategy to the equipment system or a control system of the equipment system so that the equipment system can set control parameters and work based on the target control strategy;

and/or acquiring the actual indoor environment of the building and the energy consumption of the actual equipment system under the target control strategy, and updating and training the optimization control strategy model according to the actual indoor environment of the building and the energy consumption of the actual equipment system.

9. A display platform, wherein the display platform displays an optimized control interface of an equipment system, the optimized control interface displays a target control strategy of the equipment system, and the target control strategy comprises target values of a plurality of control parameters of the equipment system; the optimization control interface displays the current values of the control parameters;

10. A cloud server, comprising: at least one memory and at least one processor; the memory stores one or more computer-executable instructions that are invoked by the processor to perform a control optimization method for an apparatus according to any one of claims 1-8.

11. A storage medium, wherein the storage medium may store one or more computer-executable instructions that, when executed, implement a method of control optimization for an apparatus of any of claims 1-8.