Heat supply system heat load prediction device and method based on TVF-EMD + PSO + SBiLSTM

Abstract

The invention discloses a heat supply system thermal load prediction device and method based on TVF-EMD + PSO + SBiLSTM, and the device comprises a data collection module which is used for obtaining the related operation data of a heat supply system; the TVF-EMD module is used for generating new data features through empirical mode decomposition of a time-varying filter according to the related operation data; the data set construction module is used for fusing the new data features and the predicted weather data and constructing a thermal load prediction data set in combination with the prediction features; the prediction model construction module is used for constructing a prediction model according to the stacked bidirectional long-short-term memory neural network and the thermal load prediction data set; and the thermal load prediction module is used for performing prediction through the prediction model to obtain a thermal load prediction result. According to the method, short-term prediction of the thermal load of the heat supply system is realized, and the precision and accuracy of thermal load prediction are effectively improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of heating systems, and in particular relates to a device and method for predicting heat load of a heating system based on TVF-EMD+PSO+SBiLSTM. Background Art

[0002] The stable operation of the heating system depends on the high accuracy and efficiency of short-term heat load forecasting. Accurate load forecasting is a key factor in ensuring that the heating system can operate efficiently and energy-efficiently. Therefore, in-depth understanding and analysis of the inherent characteristics of heating system operation data and improving forecast accuracy have become a challenging and extremely important topic in the field. Summary of the invention

[0003] In order to solve the above technical problems, the present invention proposes a device and method for predicting heat load of a heating system based on TVF-EMD+PSO+SBiLSTM to solve the problems existing in the above-mentioned prior art.

[0004] To achieve the above object, the present invention provides a heat load prediction device for a heating system based on TVF-EMD+PSO+SBiLSTM, comprising:

[0005] Data acquisition module, used to obtain relevant operation data of the heating system;

[0006] A TVF-EMD module, for generating new data features through empirical mode decomposition of a time-varying filter according to the relevant operation data;

[0007] A data set construction module, used to fuse the new data features and the forecasted weather data, and to construct a heat load forecast data set in combination with the forecast features;

[0008] A prediction model building module for building a prediction model based on a stacked bidirectional long short-term memory neural network and a heat load prediction dataset;

[0009] The heat load prediction module is used to perform prediction through the prediction model to obtain a heat load prediction result.

[0010] Preferably, the relevant operating data include: a transient power per hour, a transient flow rate, a primary water supply temperature, a primary return water temperature, a secondary water supply temperature, a secondary return water temperature, an indoor temperature, a maximum outdoor temperature, a minimum outdoor temperature, an outdoor humidity, an outdoor wind speed and a solar radiation value in the heating system.

[0011] Preferably, the TVF-EMD module includes a modal decomposition unit and a parameter optimization unit;

[0012] The modal decomposition unit is used to process the relevant operation data features per hour using the empirical mode decomposition of the time-varying filter to obtain an intrinsic mode function with a center frequency and a bandwidth;

[0013] The parameter optimization unit is used to optimize the number of intrinsic mode functions and regularization parameters in TVF-EMD based on a particle swarm optimization method.

[0014] Preferably, the stacked bidirectional long short-term memory neural network is a two-layer bidirectional long short-term memory network stack.

[0015] Preferably, the data set construction module includes:

[0016] An input feature construction unit, used to fuse the new data features with the weather data forecasted for the next hour to obtain input features;

[0017] A prediction feature building unit is used to use the instantaneous power in the next hour as a prediction feature;

[0018] The data set construction unit is used to integrate the input features and the prediction features to construct a heat load prediction data set; and divide the heat load prediction data set into a training set, a validation set and a test set.

[0019] Preferably, the device further comprises a preprocessing module, and the preprocessing module is used to perform data preprocessing on the collected heating system data;

[0020] The preprocessing module comprises:

[0021] A deleting unit, used to delete duplicate data in the relevant operation data;

[0022] A filling unit, used to fill missing data in the relevant operation data by using a forward filling method;

[0023] The outlier detection unit is used to judge and process outliers of the relevant operating data using a 3-sigma method.

[0024] Preferably, the device further comprises a heating system control module for controlling the heating system based on the heat load prediction result.

[0025] The present invention also discloses a method for predicting heat load of a heating system based on TVF-EMD+PSO+SBiLSTM, comprising the following steps:

[0026] Obtain relevant operating data of the heating system;

[0027] According to the relevant operation data, new data features are generated by empirical mode decomposition of a time-varying filter;

[0028] fusing the new data features with the forecasted weather data, and constructing a heat load forecasting data set in combination with the forecast features;

[0029] Building a prediction model based on a stacked bidirectional long short-term memory neural network and a heat load prediction dataset;

[0030] The prediction model is used to perform prediction and obtain a heat load prediction result.

[0031] The present invention also discloses a heat load prediction system for a heating system based on TVF-EMD+PSO+SBiLSTM, comprising: a memory and a processor, wherein a computer program executed by the processor is stored on the memory, and the computer program executes the above method when executed by the processor.

[0032] The present invention also discloses a storage medium, on which a computer program is stored, and the computer program executes the method when running.

[0033] Compared with the prior art, the present invention has the following advantages and technical effects:

[0034] The invention discloses a device and method for predicting heat load of a heating system based on TVF-EMD+PSO+SBiLSTM, comprising: a data acquisition module, used for acquiring relevant operation data of a heating system; a TVF-EMD module, used for generating new data features through empirical mode decomposition of a time-varying filter according to the relevant operation data; a data set construction module, used for fusing new data features with forecasted weather data, and constructing a heat load prediction data set in combination with the prediction features; a prediction model construction module, used for constructing a prediction model according to a stacked bidirectional long short-term memory neural network and a heat load prediction data set; and a heat load prediction module, used for performing prediction through the prediction model to obtain a heat load prediction result.

[0035] The present invention performs empirical mode decomposition of a time-varying filter on heating heat load data to generate new data features. Compared with the traditional method of using directly collected data to predict heat load, the present invention integrates the new data features generated by TVF-EMD and weather forecast data, constructs a new data set, and improves the quality of heat load prediction data. Compared with simple machine learning and deep learning models, the present invention proposes a prediction model based on a stacked bidirectional long short-term memory network module. The stacked bidirectional long short-term memory network considers information from forward and backward sequences at the same time to learn complex relationships in time series, can effectively process time series data, and improves the comprehensive performance of the model. Based on the fused multi-feature heat load prediction data set and the prediction model based on SBiLSTM, the heat load prediction of the heating system is performed to obtain the result of the heat load prediction. The present invention can improve the accuracy of the short-term heat load prediction of the heating system and achieve the goal of reducing energy waste. The heating system controls the corresponding hardware through the heat load value output by the prediction model to achieve efficient energy management. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0037] Figure 1 is a schematic diagram of the structure of a device according to an embodiment of the present invention;

[0038] Figure 2 Schematic diagram of the prediction method structure of an embodiment of the present invention. DETAILED DESCRIPTION

[0039] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0040] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0041] Embodiment 1:

[0042] like Figure 1 As shown, the embodiment of the present invention provides a heat load prediction device for a heating system based on TVF-EMD+PSO+SBiLSTM, comprising:

[0043] Data acquisition module, used to obtain relevant operation data of the heating system;

[0044] The collected heating system-related operating data include primary transient power, primary transient flow, primary water supply temperature, primary return water temperature, secondary water supply temperature, secondary return water temperature, indoor temperature, outdoor temperature - maximum temperature, outdoor temperature - minimum temperature, outdoor humidity, outdoor wind speed and solar radiation value per hour.

[0045] The TVF-EMD module is used to generate new data features through empirical mode decomposition of time-varying filters based on relevant operating data;

[0046] The TVF-EMD module includes:

[0047] New feature units are generated, and the hourly transient power, primary transient flow, primary water supply temperature, primary return water temperature, secondary water supply temperature, secondary return water temperature, indoor temperature, outdoor temperature-maximum temperature, outdoor temperature-minimum temperature, outdoor humidity, outdoor wind speed and solar radiation value data are processed using the empirical mode decomposition of time-varying filters (TVF-EMD) to obtain a series of intrinsic mode functions (IMFs) with specific center frequencies and bandwidths.

[0048] The parameter optimization unit is used to optimize the number k of each characteristic IMF and the regularization parameter α in the TVF-EMD using the particle swarm optimization method (PSO) to obtain the optimal number k of IMF and the regularization parameter α for each characteristic.

[0049] TVF-EMD is an improved empirical mode decomposition (EMD) method that introduces a time-varying filter to solve the problems encountered by traditional EMD when dealing with nonlinear and non-stationary signals. TVF-EMD includes the following steps:

[0050] TVF denoising, first of all, the original signal x t Use TVF to remove trend terms and high-frequency noise to get y t , the calculation formula is as follows:

[0051]

[0052] Among them, α is the regularization parameter, which controls the balance between the total variation term and the signal fidelity term. A larger α will make the denoising effect stronger, but may over-smooth the signal.

[0053] EMD decomposition, the denoised signal y t Apply EMD to decompose it into multiple IMFs and a residual signal.

[0054]

[0055] Among them, IMF k is the kth IMF, and r is the residual signal.

[0056] The particle swarm optimization method (PSO) is used to optimize the number k of optimal IMFs for each feature of TVF-EMD and the regularization parameter α.

[0057] The steps of the particle swarm optimization method include:

[0058] Initialization, particle swarm initialization, randomly generate a group of particles, each particle is a parameter combination in TVF-EMD (the number of IMFs k and the regularization parameter α), where the value range of k is 2-20, the value range of α is 0.01-0.5, and the number of particle swarms is 20.

[0059] Define the objective function and use envelope entropy as the objective function to evaluate the performance of each particle. The minimum envelope entropy indicates that the found parameters are optimal.

[0060] Update particles and update the speed and position of each particle according to the rules of the PSO algorithm. The new position of the particle is determined by the current position, the individual historical best position (pbest) and the global historical best position (gbest).

[0061] Iterative optimization, repeat the above steps, when the preset stop condition is met, that is, when the maximum number of iterations is 500, stop optimizing the number of IMFs k and the regularization parameter α.

[0062] The optimal result output is the number k of the optimal IMF for each feature, that is, k 一次瞬态功率 , k 一次瞬态流量 , k 一次供水温度 , k 一次回水温度 , k 二次供水温度 , k 二次回水温度 , k 室内温度 , k 室外温度-最高温 , k 室外温度-最低温 , k 室外湿度 , k 室外风速 , k 太阳辐射值 .

[0063] The dataset construction module is used to fuse new data features with forecasted weather data and construct a heat load forecast dataset in combination with the forecast features;

[0064] The dataset construction module includes an input feature construction unit, which is used to decompose and calculate the IMF1, IMF2…IMF1 of the first 48 hours of TVF-EMD data on primary transient power, primary transient flow, primary water supply temperature, primary return water temperature, secondary water supply temperature, secondary return water temperature, indoor temperature, outdoor temperature-highest temperature, outdoor temperature-lowest temperature, outdoor humidity, outdoor wind speed and solar radiation value. kThe fusion is performed, and the outdoor temperature-maximum temperature forecast for the next hour, the outdoor temperature-minimum temperature forecast for the next hour, the outdoor humidity forecast for the next hour, the outdoor wind speed forecast for the next hour, and the solar radiation data forecast for the next hour are added as the input features of the data set.

[0065] A prediction feature building unit is used to use the instantaneous power in the next hour as the prediction feature of the data set;

[0066] The data set construction unit is used to integrate input features and prediction features to construct a heat load prediction data set; and divide the heat load prediction data set into a training set, a validation set, and a test set.

[0067] The constructed multi-feature fused heat load dataset of the heating system is randomly shuffled, 80% of the data is used as the training and validation sets, and the remaining 20% ​​of the data is used as the test dataset.

[0068] like Figure 2 As shown, a prediction model building module is used to build a prediction model based on a stacked bidirectional long short-term memory neural network and a heat load prediction data set;

[0069] In this embodiment, a stacked bidirectional long short-term memory network (SBiLSTM) module is formed by stacking two layers of bidirectional long short-term memory networks (BiLSTM). For each training sample, its data dimension is k 一次瞬态功率 +k 一次瞬态流量 +k 一次供水温度 +k 一次回水温度 +k 二次供水温度 +k 二次回水温度 +k 室内温度 +k 室外温度-最高温 +k 室外温度-最低温 +k 室外湿度 +k 室外风速 +k 太阳辐射值 +5. Input the training samples into SBiLSTM for calculation and output the calculation results.

[0070] It includes a model performance evaluation unit, which uses the mean absolute percentage error to evaluate the performance of the integrated learning model in the training phase and the testing phase to obtain evaluation results.

[0071] In this embodiment, MAPE (Mean Absolute Percentage Error) is used to evaluate the performance of the model in both the training and testing stages of the prediction model. The calculation formula of MAPE is as follows:

[0072]

[0073] Among them, y iis the actual value, is the predicted value, and n is the number of samples.

[0074] MAPE is an intuitive measure of model accuracy, ranging from 0 to 1. The closer the value is to 0, the higher the prediction accuracy of the model.

[0075] In this embodiment, the prediction model is trained with the training data set, and the performance of the prediction model is tested with the test set, and finally the MAPE value of the prediction model under the test set is output.

[0076] As an implementation of an embodiment of the present invention, a heating system control module is further included, which is used to control the heating system based on the result of the heat load prediction. In this embodiment, the actual heating power per square meter obtained by dividing the instantaneous power predicted in the next hour by the total actual heating area can be used as the predicted heating amount of the heat source, thereby effectively controlling the heat source part of the heating system and reducing the energy consumption of the heat source.

[0077] The heat load prediction module is used to perform prediction through the prediction model to obtain the heat load prediction result.

[0078] Embodiment 2:

[0079] The embodiment of the present invention also provides a method for predicting heat load of a heating system based on TVF-EMD+PSO+SBiLSTM, comprising:

[0080] Step S1, collecting operation data of the heating system;

[0081] Step S2, preprocessing the collected heating system data;

[0082] Step S3, processing the preprocessed heating system data, and generating new data features through empirical mode decomposition of a time-varying filter;

[0083] Step S4, integrating the empirical mode decomposition through the time-varying filter to generate new data features and constructing a heat load prediction data set;

[0084] Step S5, constructing a SBiLSTM model using a stacked bidirectional long short-term memory network model;

[0085] Step S6: Based on the constructed SBiLSTM model and the heat load prediction data set, the heat load of the heating system is predicted to obtain a heat load prediction result.

[0086] The preprocessing includes: deleting duplicate data in the relevant operation data;

[0087] Filling missing data in the relevant operation data using forward filling method;

[0088] The 3-sigma method is used to judge and process abnormal values ​​of the relevant operating data.

[0089] As an implementation of the embodiment of the present invention, step S3 includes:

[0090] Step S31, performing TVF-EMD decomposition on the pre-processed data of the heating system to generate new data features;

[0091] Step S32, optimizing the parameters of TVF-EMD by particle swarm optimization method to obtain the optimal number of IMFs and regularization parameter values;

[0092] As an implementation method of an embodiment of the present invention, the stacked bidirectional long short-term memory network is formed by stacking two layers of bidirectional long short-term memory networks. For each training sample, its data dimension is k 一次瞬态功率 +k 一次瞬态流量 +k 一次供水温度 +k 一次回水温度 +k 二次供水温度 +

[0093] k 二次回水温度 +k 室内温度 +k 室外温度-最高温 +k 室外温度-最低温 +k 室外湿度 +k 室外风速 +k 太阳辐射值 +5, input the training sample into SBiLSTM for calculation and output the final calculation result.

[0094] As an implementation of an embodiment of the present invention, constructing a heat load prediction data set in step S4 includes:

[0095] The new data features are integrated with the weather data predicted for the next hour to obtain input features;

[0096] The transient power in the next hour is used as the prediction feature;

[0097] The input features and the prediction features are integrated to construct a heat load prediction data set; and the heat load prediction data set is divided into a training set, a validation set and a test set.

[0098] Embodiment three:

[0099] An embodiment of the present invention also provides a heating system heat load prediction system based on TVF-EMD+PSO+SBiLSTM, comprising: a memory and a processor, wherein the memory stores a computer program run by the processor, and the computer program executes a heating system heat load prediction method when run by the processor.

[0100] Embodiment 4:

[0101] An embodiment of the present invention further provides a storage medium, on which a computer program is stored, and the computer program executes a method for predicting heat load of a heating system when running.

[0102] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A heat load prediction device for a heating system based on TVF-EMD+PSO+SBiLSTM, characterized in that: include: Data acquisition module, used to obtain relevant operation data of the heating system; A TVF-EMD module, for generating new data features through empirical mode decomposition of a time-varying filter according to the relevant operation data; A data set construction module, used to fuse the new data features and the forecasted weather data, and to construct a heat load forecast data set in combination with the forecast features; A prediction model building module for building a prediction model based on a stacked bidirectional long short-term memory neural network and a heat load prediction dataset; The heat load prediction module is used to perform prediction through the prediction model to obtain a heat load prediction result.

2. The device according to claim 1, characterized in that The relevant operating data include: primary transient power, primary transient flow, primary water supply temperature, primary return water temperature, secondary water supply temperature, secondary return water temperature, indoor temperature, maximum outdoor temperature, minimum outdoor temperature, outdoor humidity, outdoor wind speed and solar radiation value in the heating system every hour.

3. The device according to claim 1, characterized in that The TVF-EMD module includes a modal decomposition unit and a parameter optimization unit; The modal decomposition unit is used to process the relevant operation data features per hour using the empirical mode decomposition of the time-varying filter to obtain an intrinsic mode function with a center frequency and a bandwidth; The parameter optimization unit is used to optimize the number of intrinsic mode functions and regularization parameters in TVF-EMD based on a particle swarm optimization method.

4. The device according to claim 1, characterized in that The stacked bidirectional long short-term memory neural network is a two-layer bidirectional long short-term memory network stack.

5. The device according to claim 1, characterized in that The data set construction module includes: An input feature construction unit, used to fuse the new data features with the weather data forecasted for the next hour to obtain input features; A prediction feature building unit is used to use the instantaneous power in the next hour as a prediction feature; The data set construction unit is used to integrate the input features and the prediction features to construct a heat load prediction data set; and divide the heat load prediction data set into a training set, a validation set and a test set.

6. The device according to claim 1, characterized in that The device also includes a preprocessing module, which is used to perform data preprocessing on the collected heating system data; The preprocessing module comprises: A deleting unit, used to delete duplicate data in the relevant operation data; A filling unit, used to fill missing data in the relevant operation data by using a forward filling method; The outlier detection unit is used to judge and process outliers of the relevant operating data using a 3-sigma method.

7. The device according to claim 1, characterized in that The device also includes a heating system control module, which is used to control the heating system based on the heat load prediction result.

8. A method for predicting heat load of a heating system based on TVF-EMD+PSO+SBiLSTM, characterized in that: The following steps are involved: Obtain relevant operating data of the heating system; According to the relevant operation data, new data features are generated by empirical mode decomposition of a time-varying filter; fusing the new data features with the forecasted weather data, and constructing a heat load forecasting data set in combination with the forecast features; Building a prediction model based on a stacked bidirectional long short-term memory neural network and a heat load prediction dataset; The prediction model is used to perform prediction and obtain a heat load prediction result.

9. A heat load prediction system for heating system based on TVF-EMD+PSO+SBiLSTM, characterized in that: include: A memory and a processor, wherein the memory stores a computer program executed by the processor, and when the computer program is executed by the processor, the method according to claim 8 is performed.

10. A storage medium, characterized in that: The storage medium stores a computer program, which executes the method according to claim 8 when running.

Images