CN118131642A - Control method and device of cold and hot source water system, storage medium and electronic device - Google Patents

Abstract

The application discloses a control method and device of a cold and hot source water system, a storage medium and an electronic device. Wherein the method comprises the following steps: generating time sequence data by utilizing historical user data, historical weather data and running states of a cold and hot source water system of cell users in a target cell; training the original neural network model by using the time sequence data to obtain a target neural network model; acquiring current user data, current weather data and current running state of a cold and hot source water system of a cell user of a target cell; and inputting a target neural network model to obtain a control instruction of the cold and hot source water system so as to control the running state of the cold and hot source water system according to the requirement, wherein the target neural network model is also used for recording newly received second requirement data which is in an effective period in the running process. The application solves the technical problem that the cold and hot source water system is not energy-saving enough in the running process.

Description

Control method and device of cold and hot source water system, storage medium and electronic device

Technical Field

The application relates to the technical field of building automatic control, in particular to a control method and device of a cold and hot source water system, a storage medium and an electronic device.

Background

This section is intended to provide a background or context for the matter recited in the claims or specification, which is not admitted to be prior art by inclusion in this section.

The cold and hot source water system refers to a centralized energy system for providing cold and hot water supply for buildings or facilities, and generally comprises a cold source water system and a heat source water system which are respectively responsible for cooling in summer and heating in winter, and supplying domestic hot water all the year round.

In the refrigeration mode, the water chilling unit operates, and the refrigerant absorbs heat and evaporates in the evaporator to cool the chilled water; chilled water is conveyed to the tail end of the indoor air conditioner through a pipeline, and flows back to the water chilling unit after absorbing indoor heat, and the circulation is repeated. After absorbing the heat of the refrigerant in the condenser, the cooling water radiates heat through the cooling tower and then returns to the condenser. When in a heating mode, a hot water unit operates to generate hot water; the hot water is conveyed to the heat dissipation terminal through the pipeline, and flows back to the hot water unit after releasing heat indoors, and the circulation is repeated.

The cold and hot source water system realizes unified management and efficient operation of air conditioning, heating and hot water demands in a building by intensively supplying cold and hot water, has the advantages of energy conservation, convenient maintenance, convenient realization of intelligent control and the like, and is widely applied to occasions such as large public buildings, commercial complexes, residential communities and the like. However, the cold and hot source water system always operates according to a specific operation mode in the heating or cooling process, and the actual requirements of users are not considered, so that energy conservation is not achieved.

In view of the above problems, no effective solution has been proposed at present.

Disclosure of Invention

The embodiment of the application provides a control method and device of a cold and hot source water system, a storage medium and an electronic device, which are used for at least solving the technical problem that the cold and hot source water system is not energy-saving enough in the operation process.

According to an aspect of the embodiment of the present application, there is provided a control method of a cold and hot source water system, including: generating time sequence data by utilizing historical user data, historical weather data and running states of a cold and hot source water system of cell users in a target cell; training an original neural network model by using time sequence data to obtain a target neural network model, wherein the original neural network model is used for learning the correlation of user data, weather data and the running state of a cold and hot source water system of a user in a time dimension from the time sequence data, the historical user data comprises first demand data and second demand data, the first demand data is the demand data marked as being valid all the time, the second demand data is the data marked as being valid in a specified time length, and the influence weight on an output result is larger when the time of the demand data is closer to the current time in the validity period; acquiring current user data, current weather data and current running state of a cold and hot source water system of a cell user of a target cell; the current user data, the current weather data and the current running state of the cold and hot source water system of the cell user of the target cell are input into a target neural network model to obtain a control instruction of the cold and hot source water system so as to control the running state of the cold and hot source water system according to the requirement, wherein the target neural network model is also used for recording newly received second requirement data which is in an effective period in the running process.

Optionally, generating time series data by using historical user data, historical weather data and running states of the cold and hot source water system of the cell user in the target cell includes: and taking the historical user data, the historical weather data and the running state of the cold and hot source water system on the same day as metadata, and arranging all metadata in a far-to-near date mode to obtain time sequence data.

Optionally, before generating the time series data by using the historical user data, the historical weather data and the running state of the cold and hot source water system of the cell user in the target cell, the method further comprises: the method comprises the steps of obtaining the following demand data of cell users in a target cell: the conventional leaving time, the conventional returning time and the conventional temperature provided by the user; the following associated data of cell users in a target cell are obtained: the method comprises the steps of home routing address, company routing address, home intelligent door lock routing address and cell access control routing address, wherein historical user data comprise associated data; the method comprises the steps of obtaining the following authority data of a cell user in a target cell: the method comprises the steps of reading authority information of login data in a home route, reading authority information of login data in a company route, reading authority information of door opening and closing data of an intelligent door lock, reading authority information of door entering and exiting data of a cell entrance guard and authority information of travel related data in a cell user mobile phone, wherein historical user data comprises the authority data.

Optionally, training the original neural network model by using the time sequence data to obtain a target neural network model, including: training the original neural network model by using the time sequence data to obtain an intermediate neural network model; under the condition that the identification accuracy of the intermediate neural network model reaches a preset value, taking the intermediate neural network model as a target neural network model; and under the condition that the recognition accuracy of the intermediate neural network model does not reach a preset value, training the intermediate neural network model is continued until the recognition accuracy of the intermediate neural network model reaches the preset value.

Optionally, the original neural network model includes a plurality of memory units, the plurality of memory units includes a first memory unit and a second memory unit, and in training the original neural network model using the time series data, the method includes: the first memory unit is used for recording the first demand data, and the second memory unit is used for recording the second demand data in the validity period.

Optionally, after inputting the current user data, the current weather data and the current operation state of the cold and hot source water system of the cell user of the target cell into the target neural network model, the method further comprises at least one of: under the condition that the type of the first demand data in the current user data is the same as the type of the first demand data recorded by the first memory unit, the first demand data in the current user data is saved to the first memory unit so as to cover the first demand data saved before; and under the condition that the second demand data is carried in the current user data, writing the second demand data carried in the current user data into a second memory unit.

Optionally, after recording the second demand data in the validity period by using the second memory unit, the method further includes: and deleting the second demand data exceeding the validity period from the second memory units when the second demand data in any one of the second memory units exceeds the validity period.

According to another aspect of the embodiment of the present application, there is also provided a control device for a cold and hot source water system, including: the generation unit is used for generating time sequence data by utilizing the historical user data, the historical weather data and the running state of the cold and hot source water system of the cell user in the target cell; the training unit is used for training the original neural network model by using the time sequence data to obtain a target neural network model, wherein the original neural network model is used for learning the correlation of user data, weather data and the running state of the cold and hot source water system of a user in the time dimension from the time sequence data, the historical user data comprises first demand data and second demand data, the first demand data is marked as the demand data which is valid all the time, the second demand data is marked as the data which is valid in the appointed duration, and the influence weight of the time of the demand data on the output result is larger when the time is closer to the current time in the validity period; the system comprises an acquisition unit, a control unit and a control unit, wherein the acquisition unit is used for acquiring current user data, current weather data and current running state of a cold and hot source water system of a cell user of a target cell; the control unit is used for inputting current user data, current weather data and current running state of the cold and hot source water system of the target cell into the target neural network model to obtain a control instruction of the cold and hot source water system so as to control the running state of the cold and hot source water system according to the requirement, wherein the target neural network model is also used for recording newly received second requirement data which is in an effective period in the running process.

Optionally, the generating unit is further configured to: and taking the historical user data, the historical weather data and the running state of the cold and hot source water system on the same day as metadata, and arranging all metadata in a far-to-near date mode to obtain time sequence data.

Optionally, the generating unit is further configured to: before generating time series data by utilizing historical user data, historical weather data and running states of a cold source water system and a hot source water system of a cell user in a target cell, the following demand data of the cell user in the target cell are obtained: the conventional leaving time, the conventional returning time and the conventional temperature provided by the user; the following associated data of cell users in a target cell are obtained: the method comprises the steps of home routing address, company routing address, home intelligent door lock routing address and cell access control routing address, wherein historical user data comprise associated data; the method comprises the steps of obtaining the following authority data of a cell user in a target cell: the method comprises the steps of reading authority information of login data in a home route, reading authority information of login data in a company route, reading authority information of door opening and closing data of an intelligent door lock, reading authority information of door entering and exiting data of a cell entrance guard and authority information of travel related data in a cell user mobile phone, wherein historical user data comprises the authority data.

Optionally, the training unit is further configured to: training the original neural network model by using the time sequence data to obtain an intermediate neural network model; under the condition that the identification accuracy of the intermediate neural network model reaches a preset value, taking the intermediate neural network model as a target neural network model; and under the condition that the recognition accuracy of the intermediate neural network model does not reach a preset value, training the intermediate neural network model is continued until the recognition accuracy of the intermediate neural network model reaches the preset value.

Optionally, the original neural network model includes a plurality of memory units, where the plurality of memory units includes a first memory unit and a second memory unit, and the training unit is further configured to: in the process of training the original neural network model by using the time series data, the first requirement data is recorded by using the first memory unit, and the second requirement data in the validity period is recorded by using the second memory unit.

Optionally, after inputting current user data, current weather data and current running state of the cold and hot source water system of the cell user of the target cell into the target neural network model, under the condition that the type of the first demand data in the current user data is the same as the type of the first demand data recorded by the first memory unit, storing the first demand data in the current user data into the first memory unit so as to cover the first demand data stored before; and under the condition that the second demand data is carried in the current user data, writing the second demand data carried in the current user data into a second memory unit.

Optionally, after the second demand data in the validity period is recorded by the second memory units, if the second demand data in any one of the second memory units exceeds the validity period, the second demand data exceeding the validity period is deleted from the second memory units.

According to another aspect of the embodiments of the present application, there is also provided a computer-readable storage medium including a stored program that, when run, performs the above-described method.

According to another aspect of the embodiments of the present application, there is also provided an electronic device including a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor executing the method described above by the computer program.

According to one aspect of the present application, there is provided a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer readable storage medium, and the processor executes the computer instructions to cause the computer device to perform the steps of any of the embodiments of the method described above.

In the embodiment of the application, the time sequence data is generated by utilizing the historical user data, the historical weather data and the running state of the cold and hot source water system of the cell user in the target cell; training an original neural network model by using time sequence data to obtain a target neural network model, wherein the original neural network model is used for learning the correlation of user data, weather data and the running state of a cold and hot source water system of a user in a time dimension from the time sequence data, the historical user data comprises first demand data and second demand data, the first demand data is the demand data marked as being valid all the time, the second demand data is the data marked as being valid in a specified duration, and the influence weight of the demand data on an output result is larger when the time is close to the current time in the validity period; acquiring current user data, current weather data and current running state of a cold and hot source water system of a cell user of a target cell; the method comprises the steps of inputting current user data, current weather data and current running states of cold and hot source water systems of a target cell into a target neural network model to obtain control instructions of the cold and hot source water systems so as to control the running states of the cold and hot source water systems according to requirements, wherein the target neural network model is also used for recording newly received second requirement data which are in an effective period in the running process, and further the technical problem that the cold and hot source water systems are not energy-saving enough in the running process is solved.

Drawings

The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings:

FIG. 1 is a flow chart of an alternative control method of a cold source water system according to an embodiment of the present application;

FIG. 2 is a schematic diagram of a control device of an alternative cold source water system according to an embodiment of the present application;

Fig. 3 is a block diagram of a terminal according to an embodiment of the present application.

Detailed Description

In order that those skilled in the art will better understand the present application, a technical solution in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the application described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

According to an aspect of the embodiment of the application, a method embodiment of a control method of a cold and hot source water system is provided. The requirements of users can be identified in an artificial intelligence mode, so that the load of the whole community can be counted, cooling or heating can be performed according to the requirements, the technical problem that a cold and hot source water system is not energy-saving enough in the operation process can be solved, fig. 1 is a flow chart of an alternative control method of the cold and hot source water system according to an embodiment of the application, and as shown in fig. 1, the method can comprise the following steps:

Step S1, generating time sequence data by utilizing historical user data, historical weather data and running states of a cold and hot source water system of cell users in a target cell.

The method comprises the steps of obtaining the following demand data of cell users in a target cell: the conventional leaving time, the conventional returning time and the conventional temperature provided by the user; the following associated data of cell users in a target cell are obtained: the address of home routing, the address of company routing, the routing address of home intelligent door lock and the routing address of cell entrance guard, and the historical user data comprise associated data; the method comprises the steps of obtaining the following authority data of a cell user in a target cell: the method comprises the steps of reading authority information of login data in a home route, reading authority information of login data in a company route, reading authority information of door opening and closing data of an intelligent door lock, reading authority information of door entering and exiting data of a community access control and authority information of journey related data in a community user mobile phone, wherein historical user data comprise the authority data.

In the technical scheme of the application, fields and field sequences of various data in metadata can be defined, historical user data, historical weather data and running states of a cold and hot source water system on the same day are used as metadata, and all metadata are arranged from far to near according to dates to obtain time sequence data.

And S2, training an original neural network model by using time sequence data to obtain a target neural network model, wherein the original neural network model is used for learning the correlation of user data, weather data and the running state of a cold and hot source water system in a time dimension from the time sequence data, the historical user data comprises first demand data and second demand data, the first demand data is the demand data marked as being valid all the time, the second demand data is the data marked as being valid in a specified time length, and the influence weight of the demand data on an output result is larger when the time is closer to the current time in the validity period.

In the technical scheme, training an original neural network model by using time sequence data to obtain an intermediate neural network model; under the condition that the identification accuracy of the intermediate neural network model reaches a preset value, taking the intermediate neural network model as a target neural network model; and under the condition that the recognition accuracy of the intermediate neural network model does not reach a preset value, training the intermediate neural network model is continued until the recognition accuracy of the intermediate neural network model reaches the preset value.

Optionally, the original neural network model includes a plurality of memory units, where the plurality of memory units includes a first memory unit and a second memory unit, and in training the original neural network model using the time series data, the first memory unit is used to record first demand data, and the second memory unit is used to record second demand data in a validity period.

And step S3, acquiring current user data, current weather data and current running state of the cold and hot source water system of the cell user of the target cell.

And S4, inputting current user data, current weather data and current running states of the cold and hot source water systems of the cell users of the target cell into a target neural network model to obtain control instructions of the cold and hot source water systems so as to control the running states of the cold and hot source water systems as required, wherein the target neural network model is also used for recording newly received second requirement data which are in an effective period in the running process.

After current user data, current weather data and current running state of a cold and hot source water system of a cell user of a target cell are input into a target neural network model, under the condition that the type of first demand data in the current user data is the same as the type of first demand data recorded by a first memory unit, the first demand data in the current user data is stored in the first memory unit so as to cover the first demand data stored before; under the condition that second demand data is carried in the current user data, the second demand data carried in the current user data is written into a second memory unit; and deleting the second demand data exceeding the validity period from the second memory units when the second demand data in any one of the second memory units exceeds the validity period.

As an alternative example, the following further details the technical solution of the present application in connection with the specific embodiments:

step 1, habit data (such as demand data, associated data and authority data), historical weather data, running states of a cold and hot source water system and the like of a user are obtained.

1) Providing APP or applet to the user, which provides the user with home heating habits including, but not limited to, time of departure from home, time of return to home, habit temperature, etc.; 2) Filling in a routing address in a home and a right for reading login data in a routing, a company routing address and a right for reading login data in a routing, filling in a routing address of an intelligent door lock in the home and a right for reading door opening and closing data, filling in a routing address of a community access control and a right for reading access data; 3) The user manually adjusts the data of the heating (such as adjusting the opening of a valve, adjusting the heating temperature, etc.).

It should be noted that, the habit data includes two types of long-term validity and short-term validity, after the data is filled, the storage options include "store as long-term validity", "store as this month validity", "store as this week validity", "store as this time validity", and the priority is from low to high, that is, when two habit data conflict, the priority is based on high priority, when two similar data conflict, for example, the data valid for a long time by the user is "eight out of the morning to home at eight out of the evening" and "seven out of the today to home at six in the evening" which are valid, so that the priority of the latter can be considered higher.

And 2, generating time series data of the user by using habit data of the user, wherein each metadata comprises a series of data.

And step 3, training the original long-short-term memory model by using the time sequence data to obtain a target neural network model.

And 4, performing heating control based on habit data of the user and currently input data by using the target model.

In a building self-controlled heating and cooling system, training LSTM models to optimize load control using time series based user habit data has several key benefits:

1) Accurate prediction load demand: the time sequence data records the actual demands and the change trends of users for heat supply/cold supply in different time periods, including temperature setting preference, use time period, peak load occurrence time point and the like, and the LSTM model can effectively capture the periodicity and nonlinear characteristics of user habits by utilizing the data, accurately predict the future heat/cold load demands, and is beneficial to scheduling energy supply in advance, avoiding waste caused by excessive energy supply and simultaneously preventing comfort from being reduced due to insufficient prediction.

2) And (3) personalized room temperature control: there may be significant differences in the indoor temperature requirements of different users and even the same user over different time periods. The LSTM model based on time series data is capable of learning unique lifestyle and thermal comfort preferences of each user, such as temperature setting differences between weekdays and weekends, day and night, and temporary temperature adjustment needs during special activities (e.g., family gathering, night reading, etc.). Accordingly, the building automatic control system can implement fine and personalized room temperature control, ensure that users are always in a satisfactory thermal environment, and reduce unnecessary energy consumption.

3) Dynamic response external influence: the user's demand for heating/cooling is not only influenced by his own habits, but is also closely related to external conditions such as weather changes, seasonal changes, illumination intensity, etc. The time series data contains historical information of the external factors, and the LSTM model can integrate the information, forecast the influence of the information on the load demands of users and dynamically adjust the energy supply strategy according to the information. For example, when it is predicted that sudden drop of air temperature or weakening of sunlight may cause drop of indoor temperature, the heat supply amount is increased in advance, otherwise, the heat supply amount is reduced appropriately, so that balance of supply and demand and energy saving are realized.

4) Optimizing a device operation strategy: based on the load demand predicted by the LSTM model, the building automation system may more efficiently schedule heating/cooling devices. For example, equipment maintenance, cleaning, or energy saving mode operation is scheduled based on a predicted off-peak load period, while full-load efficient operation of the equipment is ensured during a predicted peak load period. In addition, the model can also help to determine the optimal start-stop time and adjust the output power of equipment, and avoid energy waste and equipment abrasion caused by frequent start-stop.

5) Promoting energy management and energy saving target achievement: by utilizing the LSTM model trained by the time series data, the building automatic control system can more accurately match the user demands and the energy supply, reduce the non-effective energy consumption and reduce the energy waste. The energy utilization efficiency is improved, the energy conservation and emission reduction of the whole building are realized, and the requirements of green building and sustainable development are met. Meanwhile, the fine load control can also reduce the operation cost and improve the economic benefit of property management. Summarizing, the user habit data based on the time sequence is used for training the LSTM model, and for heat supply/cold supply load control of the building automatic control system, accurate prediction of load demand, personalized room temperature control, dynamic response of external influence, optimization of equipment operation strategies and effective promotion of achievement of energy management and energy saving targets can be realized, so that higher user comfort and energy efficiency are realized.

As another alternative example, the following further details the technical solution of the present application in connection with the specific embodiments:

Step 1: time series data is generated.

Step 1.1: historical data is collected.

User data acquisition, first demand data: heating demand data marked as always valid, such as constant room temperature settings, heating demand for a specific period, etc., is collected from sources of user profiles, long-term service contracts, constant lifestyle records, etc.

Second demand data: data marked as valid for a specified period of time, such as a temporary temperature adjustment request, a vacation room notification, etc., is acquired from channels such as a user temporary demand submission system, a seasonal adjustment application, a short-time special activity notification, etc., while recording the time for its effectiveness to come to rest.

Historical weather data acquisition: the calendar year weather data of the area where the target cell is located is downloaded from the approaches of a right weather service organization, a public weather data interface and the like, and the calendar year weather data comprises basic weather parameters such as air temperature, humidity, wind speed, wind direction, sunlight intensity and the like every day/hour. For special weather events (such as cold tide early warning, frost early warning, extreme high temperature, etc.) which may affect heating requirements, the special weather events can be obtained from professional weather reports and early warning systems and combined with basic weather data.

And acquiring running state data of the cold and hot source water system: historical operation records including key parameters such as boiler water outlet temperature, backwater temperature, circulation flow, combustion power, start-stop time, fault alarm and the like are extracted from a Building Automation System (BAS), an Energy Management System (EMS) or other equipment monitoring platforms.

Step 1.2: and (5) preprocessing data.

Data cleaning: missing values are checked and processed, and missing weather or system operation data can be filled in by interpolation methods (such as linear interpolation, nearest neighbor filling and the like). The removal of significant outliers, such as temperature readings outside of reasonable ranges, unreasonable operating conditions, etc., can be identified by statistical analysis (e.g., quartile law) or by thresholding based on domain knowledge. Data normalization/normalization: the numerical data (such as air temperature, humidity, power and the like) are standardized or normalized, so that the characteristics of different dimensions and different numerical ranges are ensured to be on the same scale, and the neural network training is facilitated. And (3) processing second demand data: and associating the second demand data with the time stamp corresponding to the effective time period of the second demand data to form a 'demand event' sequence with a time label. For an effective demand across multiple time steps, it is broken down into multiple single step effective demand events, each event corresponding to its exact position in the time sequence.

Step 1.3: a time series is constructed.

And (3) time sequence construction: and integrating the cleaned and preprocessed user data (the first demand data and the second demand event sequence), the weather data and the cold and hot source water system running state data according to a time sequence to form a multivariable time sequence data set taking time as an index.

Sequence division: the time series is divided into training sets, validation sets and test sets (e.g., 80%, 10% in time order) according to training requirements, ensuring spatio-temporal continuity of the data sets.

Step 2: and training a target neural network model.

Step 2.1: selecting and designing a model architecture.

Selecting a model type: deep learning models suitable for processing time series data, such as long and short term memory networks (LSTM), gate loop units (GRUs), convolutional loop neural networks (ConvLSTM), and the like, are selected.

Defining a model structure: input layer: input layers capable of receiving multivariate time series characteristics, such as multi-channel inputs (user data, weather data, system state data each as independent channels) are designed.

And (3) a circulating layer: and configuring a proper circulating neural network layer, and capturing long-term dependency relationship of time series data.

Weight distribution layer: for the second demand data, a weight distribution mechanism is designed, such as a time sensitive attention layer or adding a time decay term in the loss function.

Output layer: the output layer is set to be a structure capable of predicting the control instruction of the cold and hot source water system at the next moment, such as directly predicting control parameters (temperature and flow) or probability distribution (such as for a multi-stage control strategy).

Step 2.2: model training and tuning.

Initializing parameters and super parameters: initial model parameters (e.g., weights, offsets) are set and super parameters such as learning rate, optimizer type (e.g., adam), loss function (e.g., mean square error), batch size, sequence length, etc. are determined.

Model training: the model is gradient descent trained using the training set data, model parameters are updated by back propagation, minimizing the loss function.

Model verification and adjustment: model performance, such as prediction accuracy, MAE, RMSE, etc., is evaluated on the validation set. And adjusting super parameters or model structures according to the verification result, such as adjusting learning rate, increasing hidden layer node number, changing circulating layer type and the like.

Model test: and (3) carrying out comprehensive evaluation on the test set to ensure that the generalization capability of the model is good and no over-fitting or under-fitting phenomenon exists.

Step 3: and predicting in real time and controlling according to the requirement.

And acquiring real-time data, monitoring whether the first demand data has variation in real time, and updating to the latest state if the first demand data has variation. And monitoring a second demand data submitting interface, and timely acquiring the newly generated effective demand and the effective period thereof.

And (3) collecting real-time weather data: and the weather conditions of the current and future time periods of the cell are automatically updated by subscribing the real-time or short-time forecast data stream provided by the authoritative weather service.

And (3) monitoring the state of a cold and hot source water system: the running state parameters (such as outlet water temperature, return water temperature, circulation flow, combustion power and the like) of the current system are continuously acquired through a building automation system or equipment monitoring platform.

Model reasoning and instruction generation, and model input preparation: integrating the user data, weather data and system state data acquired in real time into a multivariate time series data point matched with the input format of the model. Model reasoning: and inputting the prepared data into a trained target neural network model, and outputting a cold and hot source water system control instruction at the next moment by the model.

And the instruction execution and feedback are carried out, a control instruction predicted by the model is sent to a controller of the cold and hot source water system, and the controller adjusts the equipment operation parameters according to the control instruction.

System response monitoring and feedback: and the response condition of the system to the control instruction is monitored in real time, and the outlet water temperature, the return water temperature, the circulation flow and the like are actually achieved. Actual operational data is fed back to the model for online learning and self-correction of the model (e.g., when a reinforcement learning framework is employed).

By adopting the technical scheme of the application, the fine prediction and control can be realized: the complex nonlinear relation is captured through the deep learning model, so that the refined prediction of the system control instruction is realized, the heat is supplied according to the need, the comfort level of a user is improved, and the energy consumption is reduced. Dynamic response and personalization services: the model can quickly respond to the short-term and temporary demand change of the user, adapt to the living habit and preference of the user and provide personalized heating service. Real-time adaptation and self-optimization: the model receives newly generated second demand data and real-time weather data in real time, continuously learns and optimizes, and keeps synchronous with the environmental conditions and the demands of users. Resource saving and environmental protection: by accurately matching the heat supply requirement, invalid heat supply and excessive heat supply are avoided, the energy consumption is obviously reduced, and the energy-saving and carbon-reducing targets are met.

In an alternative embodiment, the piping directly from the gas fired boiler to each building is a main piping, the piping from the building to each unit floor is a branch piping, and the piping in each home (or room) is an end piping. For a pipe (branch pipe or end pipe, hereinafter, a branch pipe is taken as an example for illustration, and the end pipe can be replaced in practice) with a fault, the pipe can be dredged by pressurizing, and in the case that the artificial intelligence model identifies that a specific branch pipe is blocked (such as a blockage caused by rust, a blockage caused by a falling part, and the like) by using the acquired infrared image, the dredging is performed as follows: firstly, determining the blockage level, obtaining a pressure value 1 at a valve (outflow) before flowing into a blockage position and a pressure value 2 at the valve (inflow) before flowing out of the blockage position, and obtaining the current blockage levelMathematical symbol/>Representing a downward rounding; the number f (p) of other branch pipes needing to be closed is determined according to the current blocking level (the higher the current blocking level p is, the more severe the blocking is, the more the number of other branch pipes needing to be closed is)Wherein x represents the total number of other branch pipes, is a fixed value, q is the highest blockage level allowed by the measure, is a fixed value, such as 10 (for some abnormally serious blockage, dredging by the method is impossible), and p is the actual blockage level, and can be greater than q, equal to q or less than q, and takes a natural value; and closing the other branch pipeline according to the determined quantity f (p), so that the blockage is dispersed and washed away by using high-pressure water flow in a pressurizing mode, and the pipeline is recovered to be normal.

In addition, can design two spaces about boiler inside is (specifically realize through two-layer baffle, baffle A and baffle B of below, and the same position of both sets up the limbers, and baffle A can rotate to can realize opening, partial closing, the whole closing of limbers on the baffle B), boiler lower part space direct heating, upper portion indirect heating and allow two space water to mix the heat supply, this scheme possesses the advantage in the following energy-conservation and the flexibility aspect: 1) Layered heating improves the heat efficiency, and heat energy cascade utilization: the water in the lower space directly contacts the heat source to obtain higher temperature, so that the high-efficiency utilization of heat energy is realized; the water in the upper space absorbs heat from the lower space through heat conduction to form a temperature gradient, so that the secondary utilization of heat energy is realized, the heat loss is reduced by the design, and the overall heat efficiency of the boiler is improved. 2) Accurate temperature control meets the differentiation requirement, and flexible proportioning is realized, so that the device is suitable for various loads: by controlling the mixing proportion of the water bodies in the upper space and the lower space, the proper water supply temperature can be provided according to different requirements of different floors or households on the heat supply temperature, and the mixed water can be supplied to the areas with conventional heat supply requirements; for floors with higher heat supply demands or households with urgent need of heating, the high-temperature water in the lower space can be independently supplied. The refined heat supply mode can better match the actual demand, and avoid overheat waste or insufficient heat supply. 3) Energy-saving operation, energy consumption reduction and energy-saving mode switching: under the condition that the whole heat supply requirement is low or only the conventional heat supply is needed in a partial area, the mixed water is mainly used for heat supply, so that the use of high-temperature water is reduced, and the energy consumption is reduced. When the temperature needs to be quickly raised or special high-temperature requirements are met, the high-temperature water in the lower space is used, so that the heat supply according to the requirements is realized, and the excessive consumption of energy sources is avoided. Heat loss is reduced: the design of the upper and lower space separation is beneficial to reducing heat transfer between the heat source and the boiler shell, especially the water temperature of the upper space is relatively low, radiation and convection heat loss to the external environment are reduced, and the overall heat efficiency of the boiler is further improved. 4) Dynamic adjustment, quick response, real-time adjustment of mixing ratio: by controlling the opening of the valve for mixing the water in the upper space and the lower space, the water supply temperature can be quickly adjusted according to the conditions of outside air temperature change, user demand change and the like, and the response speed and the adaptability of the heating system are enhanced. Emergency heat supply guarantee: when the extreme weather or sudden heat supply demand increases, the high-temperature water in the lower space can be directly led out, emergency or extra heat supply support is provided, and the stability and the reliability of a heating system are ensured. 5) The service life of equipment is prolonged, the maintenance cost is reduced, and the thermal stress is reduced: by mixing the low-temperature water and the high-temperature water, the thermal stress born by equipment such as pipelines, radiators and the like in the whole system can be effectively reduced, the ageing of materials is slowed down, the service life of the equipment is prolonged, and the maintenance and replacement cost in the long-term operation process is reduced.

It should be noted that, for simplicity of description, the foregoing method embodiments are all described as a series of acts, but it should be understood by those skilled in the art that the present application is not limited by the order of acts described, as some steps may be performed in other orders or concurrently in accordance with the present application. Further, those skilled in the art will also appreciate that the embodiments described in the specification are all preferred embodiments, and that the acts and modules referred to are not necessarily required for the present application.

From the description of the above embodiments, it will be clear to a person skilled in the art that the method according to the above embodiments may be implemented by means of software plus the necessary general hardware platform, but of course also by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present application may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) comprising instructions for causing a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to perform the method according to the embodiments of the present application.

According to another aspect of the embodiment of the present application, there is also provided a control device for a cold source water system for implementing the control method of a cold source water system. Fig. 2 is a schematic view of a control device of an alternative cold source water system according to an embodiment of the present application, and as shown in fig. 2, the device may include:

A generating unit 21, configured to generate time series data by using historical user data, historical weather data, and operation states of the cold and hot source water system of a cell user in a target cell;

The training unit 22 is configured to train the original neural network model by using the time sequence data to obtain a target neural network model, where the original neural network model is configured to learn, from the time sequence data, correlation between user data, weather data, and an operation state of the cold and hot source water system of a user in a time dimension, where the historical user data includes first demand data and second demand data, the first demand data is demand data marked as valid all the time, the second demand data is data marked as valid within a specified duration, and the closer the time is in the validity period, the greater the current impact weight on an output result in the demand data;

An obtaining unit 23, configured to obtain current user data, current weather data, and a current running state of the cold and hot source water system of a cell user of the target cell;

and the control unit 24 is configured to input current user data, current weather data, and a current operation state of the cold and hot source water system of the target cell into the target neural network model to obtain a control instruction of the cold and hot source water system, so as to control the operation state of the cold and hot source water system as required, where the target neural network model is further configured to record newly received second requirement data in an effective period in an operation process.

Optionally, the generating unit is further configured to: and taking the historical user data, the historical weather data and the running state of the cold and hot source water system on the same day as metadata, and arranging all metadata in a far-to-near date mode to obtain the time sequence data.

Optionally, the generating unit is further configured to: before generating time series data by utilizing historical user data, historical weather data and running states of a cold source water system and a hot source water system of a cell user in a target cell, acquiring the following demand data of the cell user in the target cell: the conventional leaving time, the conventional returning time and the conventional temperature provided by the user; the following associated data of the cell users in the target cell are obtained: the address of the home route, the address of the company route, the route address of the home intelligent door lock and the route address of the cell entrance guard, wherein the historical user data comprise the associated data; the following authority data of the cell users in the target cell are obtained: the method comprises the steps of reading authority information of login data in a home route, reading authority information of login data in a company route, reading authority information of door opening and closing data of an intelligent door lock, reading authority information of door entering and exiting data of a cell entrance guard and authority information of travel related data in a cell user mobile phone, wherein historical user data comprises the authority data.

Optionally, the training unit is further configured to: training an original neural network model by using the time sequence data to obtain an intermediate neural network model; taking the intermediate neural network model as the target neural network model under the condition that the identification accuracy of the intermediate neural network model reaches a preset value; and under the condition that the identification accuracy of the intermediate neural network model does not reach the preset value, training the intermediate neural network model is continued until the identification accuracy of the intermediate neural network model reaches the preset value.

Optionally, the original neural network model includes a plurality of memory units, where the plurality of memory units includes a first memory unit and a second memory unit, and the training unit is further configured to: and in the process of training the original neural network model by using the time series data, recording the first demand data by using the first memory unit, and recording the second demand data in the validity period by using the second memory unit.

Optionally, after inputting current user data, current weather data and current running state of the cold source water system of the cell user of the target cell into the target neural network model, if the type of first demand data in the current user data is the same as the type of first demand data recorded by the first memory unit, saving the first demand data in the current user data to the first memory unit so as to cover the first demand data saved before; and under the condition that the current user data carries second demand data, writing the second demand data carried in the current user data into a second memory unit.

Optionally, after the second demand data in the validity period is recorded by using the second memory units, if the second demand data in any one of the second memory units exceeds the validity period, the second demand data exceeding the validity period is deleted from the second memory units.

By the module, the time sequence data is generated by utilizing the historical user data, the historical weather data and the running state of the cold and hot source water system of the cell user in the target cell; training an original neural network model by using the time sequence data to obtain a target neural network model, wherein the original neural network model is used for learning the correlation of user data, weather data and the running state of a cold and hot source water system of a user in a time dimension from the time sequence data, the historical user data comprises first demand data and second demand data, the first demand data is marked as valid all the time, the second demand data is marked as valid data in a specified duration, and the influence weight on an output result is larger when the time is close to the current time in the valid period in the demand data; acquiring current user data, current weather data and current running state of the cold and hot source water system of a cell user of the target cell; the current user data, the current weather data of the cell users of the target cell and the current running state of the cold and hot source water system are input into the target neural network model to obtain a control instruction of the cold and hot source water system so as to control the running state of the cold and hot source water system according to the requirement, wherein the target neural network model is also used for recording newly received second requirement data which is in a valid period in the running process, and the technical problem that the cold and hot source water system is not energy-saving enough in the running process can be solved.

It should be noted that the above modules are the same as examples and application scenarios implemented by the corresponding steps, but are not limited to what is disclosed in the above embodiments. It should be noted that, the above modules may be implemented in a corresponding hardware environment as part of the apparatus, and may be implemented in software, or may be implemented in hardware, where the hardware environment includes a network environment.

According to another aspect of the embodiment of the present application, there is also provided a server or terminal for implementing the control method of the cold and hot source water system.

Fig. 3 is a block diagram of a terminal according to an embodiment of the present application, and as shown in fig. 3, the terminal may include: one or more (only one is shown in the figure) processors 301, memory 303, and transmission means 305, as shown in fig. 3, the terminal may further comprise an input output device 307.

The memory 303 may be used to store software programs and modules, such as program instructions/modules corresponding to the control method and apparatus of the cold and hot source water system in the embodiment of the present application, and the processor 301 executes the software programs and modules stored in the memory 303, thereby executing various functional applications and data processing, that is, implementing the control method of the cold and hot source water system described above. Memory 303 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, memory 303 may further include memory located remotely from processor 301, which may be connected to the terminal via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The transmission device 305 is used for receiving or transmitting data via a network, and may also be used for data transmission between the processor and the memory. Specific examples of the network described above may include wired networks and wireless networks. In one example, the transmission device 305 includes a network adapter (Network Interface Controller, NIC) that may be connected to other network devices and routers via a network cable to communicate with the internet or a local area network. In one example, the transmission device 305 is a Radio Frequency (RF) module, which is used to communicate with the internet wirelessly.

In particular, the memory 303 is used to store application programs.

The processor 301 may call the application program stored in the memory 303 through the transmission means 305 to perform the following steps:

Generating time sequence data by utilizing historical user data, historical weather data and running states of a cold and hot source water system of cell users in a target cell; training an original neural network model by using the time sequence data to obtain a target neural network model, wherein the original neural network model is used for learning the correlation of user data, weather data and the running state of a cold and hot source water system of a user in a time dimension from the time sequence data, the historical user data comprises first demand data and second demand data, the first demand data is marked as valid all the time, the second demand data is marked as valid data in a specified duration, and the influence weight on an output result is larger when the time is close to the current time in the valid period in the demand data; acquiring current user data, current weather data and current running state of the cold and hot source water system of a cell user of the target cell; and inputting current user data, current weather data and current running state of the cold and hot source water system of the cell user of the target cell into the target neural network model to obtain a control instruction of the cold and hot source water system so as to control the running state of the cold and hot source water system according to requirements, wherein the target neural network model is also used for recording newly received second requirement data which is in an effective period in the running process.

Alternatively, specific examples in this embodiment may refer to examples described in the foregoing embodiments, and this embodiment is not described herein.

It will be appreciated by those skilled in the art that the structure shown in fig. 3 is merely illustrative, and the terminal may be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a palm computer, a Mobile internet device (Mobile INTERNET DEVICES, MID), a PAD, etc. Fig. 3 is not limited to the structure of the electronic device. For example, the terminal may also include more or fewer components (e.g., network interfaces, display devices, etc.) than shown in fig. 3, or have a different configuration than shown in fig. 3.

Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be implemented by a program for instructing a terminal device to execute in association with hardware, the program may be stored in a computer readable storage medium, and the storage medium may include: flash disk, read-Only Memory (ROM), random-access Memory (Random Access Memory, RAM), magnetic disk or optical disk, etc. . .

The embodiment of the application also provides a storage medium. Alternatively, in the present embodiment, the above-described storage medium may be used to execute the program code of the control method of the cold source water system.

Alternatively, in this embodiment, the storage medium may be located on at least one network device of the plurality of network devices in the network shown in the above embodiment.

Alternatively, in the present embodiment, the storage medium is configured to store program code for performing the steps of:

Generating time sequence data by utilizing historical user data, historical weather data and running states of a cold and hot source water system of cell users in a target cell; training an original neural network model by using the time sequence data to obtain a target neural network model, wherein the original neural network model is used for learning the correlation of user data, weather data and the running state of a cold and hot source water system of a user in a time dimension from the time sequence data, the historical user data comprises first demand data and second demand data, the first demand data is marked as valid all the time, the second demand data is marked as valid data in a specified duration, and the influence weight on an output result is larger when the time is close to the current time in the valid period in the demand data; acquiring current user data, current weather data and current running state of the cold and hot source water system of a cell user of the target cell; and inputting current user data, current weather data and current running state of the cold and hot source water system of the cell user of the target cell into the target neural network model to obtain a control instruction of the cold and hot source water system so as to control the running state of the cold and hot source water system according to requirements, wherein the target neural network model is also used for recording newly received second requirement data which is in an effective period in the running process.

Alternatively, specific examples in this embodiment may refer to examples described in the foregoing embodiments, and this embodiment is not described herein.

Alternatively, in the present embodiment, the storage medium may include, but is not limited to: a usb disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a removable hard disk, a magnetic disk, or an optical disk, or other various media capable of storing program codes.

The foregoing embodiment numbers of the present application are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.

The integrated units in the above embodiments may be stored in the above-described computer-readable storage medium if implemented in the form of software functional units and sold or used as separate products. Based on such understanding, the technical solution of the present application may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing one or more computer devices (which may be personal computers, servers or network devices, etc.) to perform all or part of the steps of the method described in the embodiments of the present application.

In the foregoing embodiments of the present application, the descriptions of the embodiments are emphasized, and for a portion of this disclosure that is not described in detail in this embodiment, reference is made to the related descriptions of other embodiments.

In several embodiments provided by the present application, it should be understood that the disclosed client may be implemented in other manners. The above-described embodiments of the apparatus are merely exemplary, and the division of the units, such as the division of the units, is merely a logical function division, and may be implemented in another manner, for example, multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The foregoing is merely a preferred embodiment of the present application and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present application, which are intended to be comprehended within the scope of the present application.

Claims (10)

1. A control method of a cold and hot source water system, comprising:

Generating time sequence data by utilizing historical user data, historical weather data and running states of a cold and hot source water system of cell users in a target cell;

Training an original neural network model by using the time sequence data to obtain a target neural network model, wherein the original neural network model is used for learning the correlation of user data, weather data and the running state of a cold and hot source water system of a user in a time dimension from the time sequence data, the historical user data comprises first demand data and second demand data, the first demand data is marked as valid all the time, the second demand data is marked as valid data in a specified duration, and the influence weight on an output result is larger when the time is close to the current time in the valid period in the demand data;

acquiring current user data, current weather data and current running state of the cold and hot source water system of a cell user of the target cell;

and inputting current user data, current weather data and current running state of the cold and hot source water system of the cell user of the target cell into the target neural network model to obtain a control instruction of the cold and hot source water system so as to control the running state of the cold and hot source water system according to requirements, wherein the target neural network model is also used for recording newly received second requirement data which is in an effective period in the running process.

2. The method of claim 1, wherein generating time series data using historical user data, historical weather data, and operating conditions of the hot and cold source water system for cell users in the target cell comprises:

And taking the historical user data, the historical weather data and the running state of the cold and hot source water system on the same day as metadata, and arranging all metadata in a far-to-near date mode to obtain the time sequence data.

3. The method of claim 2, wherein prior to generating the time series data using the historical user data, the historical weather data, and the operating state of the hot and cold source water system for the cell users in the target cell, the method further comprises:

The following demand data of the cell users in the target cell are obtained: the conventional leaving time, the conventional returning time and the conventional temperature provided by the user;

The following associated data of the cell users in the target cell are obtained: the address of the home route, the address of the company route, the route address of the home intelligent door lock and the route address of the cell entrance guard, wherein the historical user data comprise the associated data;

The following authority data of the cell users in the target cell are obtained: the method comprises the steps of reading authority information of login data in a home route, reading authority information of login data in a company route, reading authority information of door opening and closing data of an intelligent door lock, reading authority information of door entering and exiting data of a community access control and authority information of journey related data in a community user mobile phone, wherein historical user data comprise the authority data.

4. The method of claim 1, wherein training the original neural network model with the time series data to obtain the target neural network model comprises:

training an original neural network model by using the time sequence data to obtain an intermediate neural network model;

taking the intermediate neural network model as the target neural network model under the condition that the identification accuracy of the intermediate neural network model reaches a preset value;

And under the condition that the identification accuracy of the intermediate neural network model does not reach the preset value, training the intermediate neural network model is continued until the identification accuracy of the intermediate neural network model reaches the preset value.

5. The method according to any one of claims 1 to 4, wherein the original neural network model includes a plurality of memory cells, the plurality of memory cells including a first memory cell and a second memory cell, and wherein during training of the original neural network model using the time series data, the method includes:

And recording the first demand data by using the first memory unit, and recording the second demand data in the validity period by using the second memory unit.

6. The method of claim 5, wherein after inputting current user data of cell users of the target cell, current weather data, and current operation state of the cold source water system into the target neural network model, the method further comprises at least one of:

Under the condition that the type of first demand data in the current user data is the same as the type of first demand data recorded by the first memory unit, storing the first demand data in the current user data into the first memory unit so as to cover the first demand data stored before;

And under the condition that the current user data carries second demand data, writing the second demand data carried in the current user data into a second memory unit.

7. The method of claim 5, wherein after recording the second demand data at a validity period with the second memory unit, the method further comprises:

and deleting the second demand data exceeding the validity period from any one of the second memory units under the condition that the second demand data exceeds the validity period.

8. A control device for a cold and hot source water system, comprising:

the generation unit is used for generating time sequence data by utilizing the historical user data, the historical weather data and the running state of the cold and hot source water system of the cell user in the target cell;

The training unit is used for training an original neural network model by using the time sequence data to obtain a target neural network model, wherein the original neural network model is used for learning the correlation of user data, weather data and the running state of a cold and hot source water system of a user in a time dimension from the time sequence data, the historical user data comprises first demand data and second demand data, the first demand data is the demand data marked as valid all the time, the second demand data is the data marked as valid in a specified duration, and the influence weight of the time, which is in the validity period and is closer to the time, on an output result is larger;

the acquisition unit is used for acquiring the current user data, the current weather data and the current running state of the cold and hot source water system of the cell user of the target cell;

The control unit is used for inputting the current user data, the current weather data and the current running state of the cold and hot source water system of the cell user of the target cell into the target neural network model to obtain a control instruction of the cold and hot source water system so as to control the running state of the cold and hot source water system according to the requirement, wherein the target neural network model is also used for recording newly received second requirement data which is in an effective period in the running process.

9. A computer readable storage medium, characterized in that the storage medium comprises a stored program, wherein the program when run performs the method of any of the preceding claims 1 to 7.

10. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor performs the method of any of the preceding claims 1 to 7 by means of the computer program.