CN110688735A - Time sequence signal trend prediction method, device, equipment and storage medium - Google Patents
Time sequence signal trend prediction method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN110688735A CN110688735A CN201910838354.6A CN201910838354A CN110688735A CN 110688735 A CN110688735 A CN 110688735A CN 201910838354 A CN201910838354 A CN 201910838354A CN 110688735 A CN110688735 A CN 110688735A
- Authority
- CN
- China
- Prior art keywords
- trend
- historical
- prediction
- time
- time sequence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Development Economics (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a time sequence signal trend prediction method. Generating a prediction model for reflecting the future trend corresponding to the historical time sequence signal by acquiring the historical trend corresponding to the historical time sequence signal, wherein the future trend is obtained by copying the historical trend, and the prediction model is specifically the corresponding relation between the time and a predicted value; and calling a prediction model, and acquiring a target prediction value corresponding to the target moment according to the corresponding relation. Therefore, the method provided by the invention can generate a corresponding prediction model for any historical time sequence signal to obtain a prediction value corresponding to the target time, completes the prediction function for any historical time sequence signal, and has universality. In addition, the time sequence signal trend prediction device, the time sequence signal trend prediction equipment and the time sequence signal trend prediction storage medium correspond to the method, and have the same beneficial effects.
Description
Technical Field
The present invention relates to the field of time series signal research and analysis, and in particular, to a time series signal trend prediction method, apparatus, device, and storage medium.
Background
The time sequence signal is widely applied to real life, such as voice, video, sales statistics, people flow statistics and the like. By predicting the trend of the time sequence signal in the future, the establishment of various strategies in reality can be guided more effectively, and a valuable reference is provided for enterprises and individuals. The time sequence signal prediction is generally divided into three parts of trend, period and noise, and the three parts are respectively predicted and then combined to form a relatively accurate prediction result. However, in many cases, it is important to predict the trend of the time-series signal only by predicting the trend change of the time-series signal enough to provide a reference for the production.
At present, a common time series signal trend prediction method is to establish a trend model base, select a suitable trend model from the trend model base by analyzing historical time series signals, and then complete trend prediction of time series signals according to the selected trend model. However, for any historical time sequence signal, only if the established trend model library is rich and perfect enough, the trend model with higher matching degree with the historical time sequence signal can be selected.
Therefore, the existing prediction method has high requirement on the completeness of a trend model library, is difficult to find all trend models with high matching degree for any historical sequence signals, and has no universality.
Disclosure of Invention
The invention aims to provide a time sequence signal trend prediction method, a time sequence signal trend prediction device, time sequence signal trend prediction equipment and a storage medium, wherein a prediction model for reflecting the future trend corresponding to a historical time sequence signal can be generated by acquiring the historical trend corresponding to the historical time sequence signal, so that a prediction value corresponding to a target moment can be acquired for any historical time sequence signal, and a prediction function is completed.
To solve the above technical problem, the present invention provides a time series signal trend prediction method, including:
acquiring a historical trend corresponding to the historical time sequence signal;
generating a prediction model for reflecting the future trend corresponding to the historical time sequence signal, wherein the future trend is obtained by copying the historical trend, and the prediction model is specifically the corresponding relation between the time and a predicted value;
and calling the prediction model, and acquiring a target prediction value corresponding to the target moment according to the corresponding relation.
Preferably, the acquiring of the historical trend corresponding to the historical time series signal specifically includes:
acquiring the historical time sequence signal;
segmenting the historical time series signal into a plurality of subsequences;
and respectively acquiring the trends of the plurality of subsequences, sequencing the trends of the subsequences according to the sequence of the subsequences in the historical time sequence signal, and forming the historical trend.
Preferably, the obtaining the trends of the plurality of sub-sequences, sorting the trends of the sub-sequences according to the sequence of the sub-sequences in the historical time-series signal, and forming the historical trend specifically includes:
respectively calculating the slope of each subsequence as the slope trend of each subsequence, sequencing each slope according to the sequence of the subsequences in the historical time sequence signal, and forming a historical slope trend;
calculating the intercept of each subsequence as the intercept trend of each subsequence, sequencing each intercept according to the sequence of the subsequences in the historical time sequence signals, and forming a historical intercept trend;
the historical slope trend and the historical intercept trend constitute the historical trend.
Preferably, the generating a prediction model for reflecting a future trend corresponding to the historical time series signal specifically includes:
determining the historical slope trend as the future slope trend of the historical time sequence signal, and generating a slope model K (t) reflecting the corresponding relation between the future time and the slope;
determining the historical intercept trend as the future intercept trend of the historical time sequence signal, and generating an intercept model B (t) reflecting the corresponding relation between the future time and the intercept;
generating the prediction model according to the slope model K (t) and the intercept model B (t):
Y(t)＝K(t)×t+B(t)
wherein t is a future time, and y (t) is a future predicted value corresponding to the future time of the time sequence signal.
Preferably, the segmenting the historical time series signal into a plurality of subsequences specifically comprises:
and equally dividing the historical time sequence signal into a plurality of subsequences according to a preset fixed interval.
Preferably, the method further comprises the following steps:
and acquiring a target time range corresponding to a preset target predicted value range according to the corresponding relation.
Preferably, the method further comprises the following steps:
acquiring an actual signal value corresponding to the target moment;
and comparing the actual signal value with a target predicted value, and adjusting the prediction model according to a comparison result.
To solve the above technical problem, the present invention further provides a time series signal trend prediction apparatus, including:
the acquisition module is used for acquiring the historical trend corresponding to the historical time sequence signal;
the generation module is used for generating a prediction model for reflecting the future trend corresponding to the historical time sequence signal, wherein the future trend is obtained by copying the historical trend, and the prediction model is specifically the corresponding relation between the time and a predicted value;
and the calling module is used for calling the prediction model and acquiring a target prediction value corresponding to the target moment according to the corresponding relation.
In order to solve the above technical problem, the present invention further provides a time series signal trend prediction device, including a memory for storing a computer program;
a processor for implementing the steps of the time series signal trend prediction method according to any one of the above when the computer program is executed.
In order to solve the above technical problem, the present invention further provides a computer-readable storage medium, having a computer program stored thereon, where the computer program is executed by a processor to implement the steps of the time series signal trend prediction method according to any one of the above items.
The invention provides a time sequence signal trend prediction method which comprises the steps of firstly, obtaining historical trends corresponding to historical time sequence signals, obtaining future trends corresponding to the historical time sequence signals by copying the historical trends, and generating a prediction model for reflecting the future trends. Aiming at different historical time sequence signals, the invention can generate a prediction model for reflecting the future trend of the historical time sequence signals. The prediction model is specifically the corresponding relation between the time and the predicted value; and acquiring a target predicted value corresponding to the target moment according to the corresponding relation by calling the prediction model. Therefore, the method provided by the invention can generate a corresponding prediction model for any historical time sequence signal to obtain a prediction value corresponding to the target time, completes the prediction function for any historical time sequence signal, and has universality.
In addition, the time sequence signal trend prediction device, the time sequence signal trend prediction equipment and the time sequence signal trend prediction storage medium correspond to the method, and have the same beneficial effects.
Drawings
In order to illustrate the embodiments of the present invention more clearly, the drawings that are needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings can be obtained by those skilled in the art without inventive effort.
Fig. 1 is a flowchart of a time series signal trend prediction method according to an embodiment of the present invention;
fig. 2 is a flowchart of a time series signal trend prediction method in an application scenario according to an embodiment of the present invention;
FIG. 3 is a block diagram of a time series signal trend prediction apparatus according to an embodiment of the present invention;
fig. 4 is a block diagram of a time-series signal trend prediction apparatus according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without any creative work belong to the protection scope of the present invention.
The core of the invention is to provide a time sequence signal trend prediction method, a time sequence signal trend prediction device, time sequence signal trend prediction equipment and a storage medium, wherein a prediction model for reflecting the future trend corresponding to a historical time sequence signal can be generated by acquiring the historical trend corresponding to the historical time sequence signal, so that the prediction value corresponding to a target moment can be acquired for any historical time sequence signal, and the prediction function is completed.
In order that those skilled in the art will better understand the disclosure, the invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
Fig. 1 is a flowchart of a time series signal trend prediction method according to an embodiment of the present invention. As shown in fig. 1, the time series signal trend prediction method includes steps S101 to S103:
step S101: acquiring a historical trend corresponding to the historical time sequence signal;
in specific implementation, a history trend corresponding to a history time sequence signal with a length of T is obtained. For example, if T of the acquired history time series signal is 10, the history trend is a history value corresponding to each of 10 times. In actual life, the historical time sequence signal can be selected as the flow of people corresponding to business hours in a day in a shopping mall or other public places, and can also be selected as noise data corresponding to 24 hours in a day. Without limitation, one skilled in the art may select the corresponding historical time-series signal according to the content predicted by the actual need.
Specifically, the acquiring of the historical trend corresponding to the historical time series signal specifically includes:
acquiring a historical time sequence signal;
segmenting the historical time sequence signal into a plurality of subsequences;
and respectively acquiring the trends of the plurality of subsequences, sequencing the trends of the subsequences according to the sequence of the subsequences in the historical time sequence signal, and forming the historical trend.
In an embodiment, if the selected time range of the historical time sequence signal is larger, the corresponding data amount is larger, in this case, the historical time sequence signal may be processed in a segmented manner, that is, the historical time sequence signal is divided into a plurality of subsequences, and the historical trends corresponding to the subsequences are respectively obtained, and then the historical trends corresponding to all the historical time sequence signals are formed by sorting the trends of the subsequences according to the sequence of the subsequences in the historical time sequence signal.
Specifically, the historical timing signal may be equally divided into a plurality of subsequences at a preset fixed interval. For example, if a fixed interval is determined to be 3 and the historical timing signals [1, 2 … … 9, 10] with a length of 10 are equally divided, the subsequences are [1, 2, 3], [4, 5, 6], [7, 8, 9], and times 3, 6, and 9 are called transition points.
Step S102: generating a prediction model for reflecting the future trend corresponding to the historical time sequence signal, wherein the future trend is obtained by copying the historical trend, and the prediction model is specifically the corresponding relation between the time and the predicted value;
step S103: and calling the prediction model, and acquiring a target prediction value corresponding to the target moment according to the corresponding relation.
In step S102, the future tendency is obtained by copying the history tendency. For example, a historical time sequence signal with the time of [1, 2 … … 9, 10] is selected, and the historical values corresponding to the 10 times are [2, 1, 6, 8, 3, 18, 14, 4, 9, 5], respectively, so that the future trend of the future time of [11, 12 … … 19, 20] corresponding to the historical time sequence signal is [2, 1, 6, 8, 3, 18, 14, 4, 9, 5], which is the same as the historical trend. In real life, if the historical time sequence signal is selected as the noise data corresponding to each time of the first day, when the noise data corresponding to the second day is predicted, the future trend corresponding to the second day can be obtained by copying the noise data of the first day. In one embodiment, if the noise data corresponding to the fifth day is predicted, the future trend corresponding to the fifth day can be obtained by copying the noise data corresponding to each time of the first day four times. Therefore, a prediction model for reflecting the future trend corresponding to the historical time-series signal can be generated according to the historical trend of the historical time-series signal.
In one embodiment, when a predicted value at a time t in the future of the historical time sequence signal needs to be obtained, the prediction model can be called, and a predicted value y (t) reflecting the future time of the historical time sequence signal can be obtained through the corresponding relation between the time and the predicted value in the prediction model.
The invention provides a time sequence signal trend prediction method which comprises the steps of firstly, obtaining historical trends corresponding to historical time sequence signals, obtaining future trends corresponding to the historical time sequence signals by copying the historical trends, and generating a prediction model for reflecting the future trends. Aiming at different historical time sequence signals, the invention can generate a prediction model for reflecting the future trend of the historical time sequence signals. The prediction model is specifically the corresponding relation between the time and the predicted value; and acquiring a target predicted value corresponding to the target moment according to the corresponding relation by calling the prediction model. Therefore, the method provided by the invention can generate a corresponding prediction model for any historical time sequence signal to obtain a prediction value corresponding to the target time, completes the prediction function for any historical time sequence signal, and has universality.
In one embodiment, the obtaining the trends of the multiple sub-sequences, and the sorting the trends of the sub-sequences according to the sequence of the sub-sequences in the historical time-series signal, and the forming the historical trend specifically includes:
respectively calculating the slope of each subsequence as the slope trend of each subsequence, sequencing each slope according to the sequence of the subsequence in the historical time sequence signal, and forming the historical slope trend;
calculating the intercept of each subsequence as the intercept trend of each subsequence, sequencing each intercept according to the sequence of the subsequence in the historical time sequence signal, and forming the historical intercept trend;
the historical slope trend and the historical intercept trend constitute a historical trend.
It can be understood that the relationship between the time of each sub-sequence and the corresponding history value can be represented by a line segment formula, and the more sub-sequences are decomposed, the more accurate the relationship between the time and the corresponding history value is represented by the line segment formula, and meanwhile, the more sub-sequences are, the larger workload is caused. Those skilled in the art can determine the specific number of sub-sequences to be decomposed according to practical situations, which is not limited in this embodiment.
Specifically, the relationship between the time of each subsequence and the corresponding historical value is represented by a line segment formula, and the historical trend can be divided into two parts, namely a historical slope trend and a historical intercept trend. And calculating the slope of each subsequence as the slope trend of each subsequence. For example, subsequences [1, 2, 3] and [4, 5, 6], where the history value for time 1 is 2, the history value for time 3 is 6, and the slope of subsequence [1, 2, 3] is determined to be (6-2)/2 ═ 2; the historical value corresponding to time 4 is 8, and the historical value corresponding to time 6 is 18, and similarly, the slope of the subsequence [4, 5, 6] is 5. Then, sequencing all slopes according to the sequence of the subsequences corresponding to the historical time sequence signals, and accordingly determining the trend of the historical slopes;
and when the historical intercept trend is calculated, the intercept of each subsequence is respectively calculated to serve as the intercept trend of each subsequence, and the intercepts are sequenced according to the sequence of the subsequences in the historical time sequence signal, so that the historical intercept trend is formed. Taking the subsequences [1, 2, 3] and [4, 5, 6] as examples, simple calculation can obtain that the intercept of the subsequences [1, 2, 3] is 0 and the intercept of the subsequences [4, 5, 6] is-12. The historical intercept trend is [0, -12] composed in the order of the subsequences in the historical timing signal.
The historical trend is divided into the historical slope trend and the historical intercept trend to be determined, so that the historical trend can be represented by a line segment formula, calculation is simpler and more convenient, and the historical trend can be obtained conveniently.
In one embodiment, a predictive model for reflecting future trends corresponding to historical timing signals may be generated by the following method.
Specifically, determining the historical slope trend as the future slope trend of the historical time sequence signal, and generating a slope model K (t) reflecting the corresponding relation between the future time and the slope; for example, at fixed interval 3, the historical time series signal is decomposed, the obtained historical slope trend is [2, 5, -2.5], the slope trend is determined as the future slope trend, and a slope model k (t) is generated to reflect the corresponding relation between the future time and the slope.
Similarly, determining the historical intercept trend as the future intercept trend of the historical time sequence signal, and generating an intercept model B (t) reflecting the corresponding relation between the future time and the intercept;
generating a prediction model according to the slope model K (t) and the intercept model B (t):
Y(t)＝K(t)×t+B(t)
wherein t is a future time, and y (t) is a future predicted value corresponding to the future time of the time sequence signal.
According to the embodiment, the influence of the historical trend on the future can be reflected more accurately by copying the historical slope trend and the historical intercept trend as the future slope trend and the future intercept trend.
Specifically, the slope may be represented by multiplying the slope variation by the transformation matrix, and the intercept may be represented by multiplying the intercept variation by the transformation matrix, so that the prediction model is:
Y(t)＝A'(t)·K'T×t+A'(t)B'T
where A ' (t) is the future transformation matrix, K ' is the future slope delta vector, and B ' is the future intercept delta vector.
Calculating a slope variation vector K ═ Δ K for each subsequence1Δk2Δk3]For each subsequence, an intercept variation vector B ═ Δ B is calculated1Δb2Δb3]. For example, [1, 2 … … 9, 10]Of the historical time sequence signal, subsequence [1, 2, 3]]、[4，5，6]And [7, 8, 9]]The slopes of (a) are 2, 5 and-2.5, respectively. The slope change amounts were 2, 3, and-7.5, respectively. In the conversion matrix A, the number of rows i is expressed as the time in the history timing signal, andthe number of columns j is denoted as change point, e.g. if the number of change points is 3, the number of columns of the conversion matrix should be 3 accordingly. A is to beijWhen 0 is expressed that the ith time is less than the change point j, A isijWhen the i-th time is equal to or greater than the j-th time, 1 is expressed. For example, for the above [1, 2 … … 9, 10]The change points of the historical time sequence signal are 3, 6 and 9
When the time t to be predicted is 15, that is, the required prediction time duration is 5, the number of the variable points needs to be expanded, the expanded variable points can be obtained by dividing the predicted time duration by a fixed interval and rounding, and in the case of the fixed interval 3, the number to be expanded is 5/3 which is equal to 1; corresponding expansion slope variation vector and intercept variation vector, where the original slope variation vector is K ═ Δ K1Δk2Δk3]The slope change vector is extended by the slope change of the replicon sequence, and K' is [ Δ K ] after the extension1Δk2Δk3Δk1]The slope variation vector is extended by the intercept variation of the replicon sequence, and the intercept variation vector is B' ═ Δ B1Δb2Δb3Δb1]The transformation matrix is expanded accordingly to:
thus, the model y (t) ═ a ' (t) · K ' is predicted 'T×t+A'(t)B'TAnd acquiring a target predicted value corresponding to the target moment. It should be noted that different prediction durations correspond to different future transformation matrices, future slope variation vectors, and future intercept variation vectors. The skilled person can correspondingly determine the corresponding parameters according to the actual required prediction duration, thereby obtaining the target prediction value at the target time.
In one embodiment, the method for predicting the trend of the time series signal further includes:
and acquiring a target time range corresponding to a preset target predicted value range according to the corresponding relation.
It can be understood that, when a target time range corresponding to the preset target predicted value range is desired to be obtained, the prediction model may be invoked, and the target time range is obtained according to the corresponding relationship therein. In practical applications, for example, in order to predict a time range of a mall with a people flow rate of over 500 people in a day, and thus arrange more staff to provide better services for customers within the time range, the time range of over 500 people can be calculated through selecting the historical people flow rate of the mall and establishing a prediction model. According to the embodiment, the target time range corresponding to the preset target prediction value range is obtained by using the prediction model, so that the prediction model is more flexibly applied, and the formulation of a strategy in real life is better guided.
In one embodiment, the method for predicting the trend of the time series signal further includes:
acquiring an actual signal value corresponding to a target moment;
and comparing the actual signal value with the target predicted value, and adjusting the prediction model according to the comparison result.
Specifically, in order to obtain a more accurate prediction model, an actual signal value corresponding to a target moment can be acquired, and simultaneously compared with a target prediction value obtained through the prediction model, so that a comparison result is analyzed, the prediction model is correspondingly adjusted, and the trend of a time sequence signal can be predicted more accurately by using the prediction model.
In order to make the technical scheme of the method clear to those skilled in the art, an application scenario is given below in which the method is applied to commodity sales statistics in business hours, and sales conditions of commodities at various times in tomorrow are predicted according to the sales conditions of the commodities at present. Fig. 2 is a flowchart of a time series signal trend prediction method in an application scenario according to an embodiment of the present invention, as shown in fig. 2, specifically including the following steps:
step S201: acquiring sales data of each moment of today as a historical time sequence signal;
step S202: taking two hours as a fixed interval, equally dividing one day into a plurality of time intervals, wherein each time interval is a subsequence;
step S203: respectively calculating the slope and the slope variation of each subsequence, and sequencing the subsequences according to the sequence of the subsequences corresponding to the historical time sequence signals to form a historical slope variation vector;
step S204: calculating the intercept and the intercept variable quantity of each subsequence respectively, and sequencing the subsequences according to the sequence of the subsequences corresponding to the historical time sequence signals to form a historical intercept variable quantity vector;
step S205: determining a future conversion matrix according to the predicted time length, and copying a historical slope variation vector as a future slope variation vector; copying a historical intercept variation vector as a future intercept variation vector;
step S206: and establishing a prediction model to predict sales data at each time in tomorrow.
It should be noted that, predicting the future trend of the historical time-series signal can only predict the future time-series signal with the same type as the historical time-series signal, for example, today's sales data is acquired in the present application scenario, and the predicted future trend can only represent the future sales data and cannot be used for representing other types of trends. The fixed interval may not be two hours, and a suitable value may be selected according to the actual situation, for example, the obtained historical timing signal has a large time range, and a large fixed interval may be selected as much as possible without affecting the prediction result in order to reduce the workload. It can be understood that, in the present embodiment, the sales data of tomorrow is predicted, and therefore, the historical slope variation vector and the historical intercept variation vector only need to be copied once, and the specific number of times of copying should be determined according to the time period to be predicted, for example, the sales data of the acquired needs to be predicted, and the historical slope variation vector and the historical intercept variation vector need to be copied twice, so as to obtain the corresponding acquired slope variation vector and acquired intercept variation vector.
The invention also provides a time sequence signal trend prediction device and a time sequence signal trend prediction device corresponding embodiment. It should be noted that the time series signal trend prediction device is based on the angle of the functional module, and the time series signal trend prediction device is based on the angle of the hardware.
FIG. 3 is a block diagram of a time series signal trend prediction apparatus according to an embodiment of the present invention; as shown in fig. 3, the present invention provides a time series signal trend prediction apparatus, which includes:
the acquisition module 10 is used for acquiring a history trend corresponding to the history time sequence signal;
the generation module 11 is configured to generate a prediction model for reflecting a future trend corresponding to the historical time sequence signal, where the future trend is obtained by copying the historical trend, and the prediction model is specifically a corresponding relationship between a time and a predicted value;
and the calling module 12 is used for calling the prediction model and obtaining a target prediction value corresponding to the target moment according to the corresponding relation.
Since the embodiments of this section correspond to the embodiments of the method section, reference is made to the description of the embodiments of the method section for the embodiments of this section, and details are not repeated here.
The invention provides a time sequence signal trend prediction device, which comprises the steps of firstly, obtaining the historical trend corresponding to a historical time sequence signal, obtaining the future trend corresponding to the historical time sequence signal by copying the historical trend, and generating a prediction model for reflecting the future trend. Aiming at different historical time sequence signals, the invention can generate a prediction model for reflecting the future trend of the historical time sequence signals. The prediction model is specifically the corresponding relation between the time and the predicted value; and acquiring a target predicted value corresponding to the target moment according to the corresponding relation by calling the prediction model. Therefore, the method provided by the invention can generate a corresponding prediction model for any historical time sequence signal to obtain a prediction value corresponding to the target time, completes the prediction function for any historical time sequence signal, and has universality.
Fig. 4 is a block diagram of a time-series signal trend prediction apparatus according to an embodiment of the present invention. As shown in fig. 4, the present invention provides a time series signal trend prediction apparatus, which includes a memory 20 for storing a computer program;
a processor 21 for implementing the steps of the time series signal trend prediction method of any one of the above when executing a computer program.
The processor 21 may include one or more processing cores, such as a 4-core processor, an 8-core processor, and the like. The processor 21 may be implemented in at least one hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), and a PLA (Programmable Logic Array). The processor 21 may also include a main processor and a coprocessor, where the main processor is a processor for processing data in an awake state, and is also called a Central Processing Unit (CPU); a coprocessor is a low power processor for processing data in a standby state. In some embodiments, the processor 21 may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content required to be displayed on the display screen. In some embodiments, the processor 21 may further include an AI (Artificial Intelligence) processor for processing a calculation operation related to machine learning.
The memory 20 may include one or more computer-readable storage media, which may be non-transitory. Memory 20 may also include high speed random access memory, as well as non-volatile memory, such as one or more magnetic disk storage devices, flash memory storage devices. In this embodiment, the memory 20 is at least used for storing the following computer program 201, wherein after being loaded and executed by the processor 21, the computer program can implement relevant steps in the time series signal trend prediction method disclosed in any of the foregoing embodiments. In addition, the resources stored in the memory 20 may also include an operating system 202, data 203, and the like, and the storage manner may be a transient storage manner or a permanent storage manner. Operating system 202 may include, among others, Windows, Unix, Linux, and the like.
In some embodiments, the server may also include an input output interface 22, a communication interface 23, a power supply 24, and a communication bus 25.
Those skilled in the art will appreciate that the configuration shown in FIG. 4 does not constitute a limitation of a time series signal trend prediction device and may include more or fewer components than those shown.
Since the embodiment of the time-series signal trend prediction device portion corresponds to the embodiment of the method portion, please refer to the description of the embodiment of the method portion for the embodiment of the time-series signal trend prediction device portion, which is not repeated here. In some embodiments of the invention, the processor and memory may be connected by a bus or other means.
The invention provides a time sequence signal trend prediction device which can realize the following method: firstly, acquiring a historical trend corresponding to a historical time sequence signal, obtaining a future trend corresponding to the historical time sequence signal by copying the historical trend, and generating a prediction model for reflecting the future trend. Aiming at different historical time sequence signals, the invention can generate a prediction model for reflecting the future trend of the historical time sequence signals. The prediction model is specifically the corresponding relation between the time and the predicted value; and acquiring a target predicted value corresponding to the target moment according to the corresponding relation by calling the prediction model. Therefore, the method provided by the invention can generate a corresponding prediction model for any historical time sequence signal to obtain a prediction value corresponding to the target time, completes the prediction function for any historical time sequence signal, and has universality.
Finally, the invention also provides a corresponding embodiment of the computer readable storage medium. The computer-readable storage medium has stored thereon a computer program which, when being executed by a processor, carries out the steps as set forth in the above-mentioned method embodiments.
It is to be understood that if the method in the above embodiments is implemented in the form of software functional units and sold or used as a stand-alone product, it can be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and performs all or part of the steps of the methods according to the embodiments of the present invention, or all or part of the technical solution. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The present invention provides a time series signal trend prediction method, apparatus, device and storage medium. The embodiments are described in a progressive manner in the specification, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description. It should be noted that, for those skilled in the art, it is possible to make various improvements and modifications to the present invention without departing from the principle of the present invention, and those improvements and modifications also fall within the scope of the claims of the present invention.
It is further noted that, in the present specification, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
Claims (10)
1. A method for predicting a trend of a time series signal, comprising:
acquiring a historical trend corresponding to the historical time sequence signal;
generating a prediction model for reflecting the future trend corresponding to the historical time sequence signal, wherein the future trend is obtained by copying the historical trend, and the prediction model is specifically the corresponding relation between the time and a predicted value;
and calling the prediction model, and acquiring a target prediction value corresponding to the target moment according to the corresponding relation.
2. The time series signal trend prediction method according to claim 1, wherein the acquiring of the historical trend corresponding to the historical time series signal specifically comprises:
acquiring the historical time sequence signal;
segmenting the historical time series signal into a plurality of subsequences;
and respectively acquiring the trends of the plurality of subsequences, sequencing the trends of the subsequences according to the sequence of the subsequences in the historical time sequence signal, and forming the historical trend.
3. The method according to claim 2, wherein the obtaining the trends of the plurality of sub-sequences, the sorting the trends of the sub-sequences according to the sequence of the sub-sequences in the historical time series signal, and the forming the historical trend specifically includes:
respectively calculating the slope of each subsequence as the slope trend of each subsequence, sequencing each slope according to the sequence of the subsequences in the historical time sequence signal, and forming a historical slope trend;
calculating the intercept of each subsequence as the intercept trend of each subsequence, sequencing each intercept according to the sequence of the subsequences in the historical time sequence signals, and forming a historical intercept trend;
the historical slope trend and the historical intercept trend constitute the historical trend.
4. The time series signal trend prediction method according to claim 3, wherein the generating of the prediction model for reflecting the future trend corresponding to the historical time series signal specifically comprises:
determining the historical slope trend as the future slope trend of the historical time sequence signal, and generating a slope model K (t) reflecting the corresponding relation between the future time and the slope;
determining the historical intercept trend as the future intercept trend of the historical time sequence signal, and generating an intercept model B (t) reflecting the corresponding relation between the future time and the intercept;
generating the prediction model according to the slope model K (t) and the intercept model B (t):
Y(t)＝K(t)×t+B(t)
wherein t is a future time, and y (t) is a future predicted value corresponding to the future time of the time sequence signal.
5. The time series signal trend prediction method according to claim 2, wherein the segmenting the historical time series signal into a plurality of subsequences specifically comprises:
and equally dividing the historical time sequence signal into a plurality of subsequences according to a preset fixed interval.
6. The time series signal trend prediction method of claim 1, further comprising:
and acquiring a target time range corresponding to a preset target predicted value range according to the corresponding relation.
7. The time series signal trend prediction method of claim 1, further comprising:
acquiring an actual signal value corresponding to the target moment;
and comparing the actual signal value with a target predicted value, and adjusting the prediction model according to a comparison result.
8. A time series signal trend prediction device, comprising:
the acquisition module is used for acquiring the historical trend corresponding to the historical time sequence signal;
the generation module is used for generating a prediction model for reflecting the future trend corresponding to the historical time sequence signal, wherein the future trend is obtained by copying the historical trend, and the prediction model is specifically the corresponding relation between the time and a predicted value;
and the calling module is used for calling the prediction model and acquiring a target prediction value corresponding to the target moment according to the corresponding relation.
9. A time series signal trend prediction apparatus, comprising a memory for storing a computer program;
a processor for implementing the steps of the time series signal trend prediction method according to any one of claims 1 to 7 when executing said computer program.
10. A computer-readable storage medium, having stored thereon a computer program which, when being executed by a processor, carries out the steps of the time-series signal trend prediction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910838354.6A CN110688735A (en) | 2019-09-05 | 2019-09-05 | Time sequence signal trend prediction method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910838354.6A CN110688735A (en) | 2019-09-05 | 2019-09-05 | Time sequence signal trend prediction method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110688735A true CN110688735A (en) | 2020-01-14 |
Family
ID=69107881
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910838354.6A Withdrawn CN110688735A (en) | 2019-09-05 | 2019-09-05 | Time sequence signal trend prediction method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110688735A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115158399A (en) * | 2022-06-14 | 2022-10-11 | 通号城市轨道交通技术有限公司 | Time sequence signal abnormity detection method and system |
| CN117689253A (en) * | 2023-12-07 | 2024-03-12 | 朴道征信有限公司 | Method, device, equipment and medium for generating flow information of product model |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105512741A (en) * | 2014-09-26 | 2016-04-20 | 山西云智慧科技股份有限公司 | Bus passenger traffic combined prediction method |
| CN106383999A (en) * | 2016-09-13 | 2017-02-08 | 北京协力筑成金融信息服务股份有限公司 | Trend analysis method and device of multi-source time sequence data |
| CN107480810A (en) * | 2017-07-21 | 2017-12-15 | 上海电力学院 | A kind of moon electricity demand forecasting method that double smoothing is improved using seasonal index number |
| CN108710757A (en) * | 2018-05-18 | 2018-10-26 | 山东大学 | Mechanical Running Condition monitoring method and device based on time-varying parameters prediction model |
| CN109948452A (en) * | 2019-02-22 | 2019-06-28 | 中国电子科技集团公司第三十六研究所 | A kind of clock signal prediction technique and device |
-
2019
- 2019-09-05 CN CN201910838354.6A patent/CN110688735A/en not_active Withdrawn
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105512741A (en) * | 2014-09-26 | 2016-04-20 | 山西云智慧科技股份有限公司 | Bus passenger traffic combined prediction method |
| CN106383999A (en) * | 2016-09-13 | 2017-02-08 | 北京协力筑成金融信息服务股份有限公司 | Trend analysis method and device of multi-source time sequence data |
| CN107480810A (en) * | 2017-07-21 | 2017-12-15 | 上海电力学院 | A kind of moon electricity demand forecasting method that double smoothing is improved using seasonal index number |
| CN108710757A (en) * | 2018-05-18 | 2018-10-26 | 山东大学 | Mechanical Running Condition monitoring method and device based on time-varying parameters prediction model |
| CN109948452A (en) * | 2019-02-22 | 2019-06-28 | 中国电子科技集团公司第三十六研究所 | A kind of clock signal prediction technique and device |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115158399A (en) * | 2022-06-14 | 2022-10-11 | 通号城市轨道交通技术有限公司 | Time sequence signal abnormity detection method and system |
| CN115158399B (en) * | 2022-06-14 | 2023-10-17 | 通号城市轨道交通技术有限公司 | Time sequence signal abnormality detection method and system |
| CN117689253A (en) * | 2023-12-07 | 2024-03-12 | 朴道征信有限公司 | Method, device, equipment and medium for generating flow information of product model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2022037337A1 (en) | Distributed training method and apparatus for machine learning model, and computer device | |
| US20180150783A1 (en) | Method and system for predicting task completion of a time period based on task completion rates and data trend of prior time periods in view of attributes of tasks using machine learning models | |
| CN112053004A (en) | Method and apparatus for time series prediction | |
| CN108052979A (en) | The method, apparatus and equipment merged to model predication value | |
| CN111883262B (en) | Epidemic situation trend prediction method and device, electronic equipment and storage medium | |
| CN106909931B (en) | Feature generation method and device for machine learning model and electronic equipment | |
| TWI526921B (en) | Method and device for displaying character string | |
| WO2021133253A1 (en) | Method and apparatus for predicting energy consumption, and device and storage medium thereof | |
| CN111898247B (en) | Landslide displacement prediction method, landslide displacement prediction equipment and storage medium | |
| CN114359563B (en) | Model training method, device, computer equipment and storage medium | |
| CN112508118A (en) | Target object behavior prediction method aiming at data migration and related equipment thereof | |
| CN110688735A (en) | Time sequence signal trend prediction method, device, equipment and storage medium | |
| CN106485348A (en) | A kind of Forecasting Methodology of transaction data and device | |
| WO2023103527A1 (en) | Access frequency prediction method and device | |
| CN112182118B (en) | Target object prediction method based on multiple data sources and related equipment thereof | |
| CN114581119A (en) | Flow prediction method and device | |
| Xu et al. | High-performance financial simulation using randomized quasi-Monte Carlo methods | |
| CN116542328B (en) | Knowledge distillation method and device for CTR prediction model | |
| CN114817845B (en) | Data processing method, device, electronic equipment and storage medium | |
| CN109933749B (en) | Method and device for generating information | |
| CN114757700A (en) | Article sales prediction model training method, article sales prediction method and apparatus | |
| CN114756352A (en) | Method, device and medium for scheduling server computing resources | |
| Syahputra et al. | Implementasi Metode Multi Attribute Utility Theory Pemberian Komisi Untuk Sales | |
| JP7378866B1 (en) | Programs, information processing systems and information processing methods | |
| CN112070607A (en) | Stock fluctuation prediction method and device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20200114 |
|
| WW01 | Invention patent application withdrawn after publication |