A day-ahead optimal scheduling method and system for a heating system considering supply and demand coordination

Abstract

The present invention discloses a method for optimizing the scheduling of a heating system in a day-ahead manner considering the coordination of supply and demand, comprising: establishing a load prediction model for each heating station on the next day, obtaining a substation demand load curve for each heating station in each time period on the next day; aligning and superimposing the demand loads of each substation in each time period on the next day according to the substation demand load curves in each time period on the next day, and establishing a total demand load curve for each time period on the next day for the heat source; modulating the total demand load curve for each time period on the next day for the heat source based on heat source constraints, heat network transmission delays, and transmission heat loss factors, and obtaining a total supply load curve for each time period on the next day for the heat source; automatically generating a substation supply load curve for each time period on the next day according to the total supply load curve for each time period on the next day for the heat source and the substation demand load curves for each heating station in each time period on the next day, and performing on-demand regulation of each heating station in each time period on the next day; constructing a day-ahead optimization scheduling plan to realize the day-ahead optimization scheduling of a heating system in a day-ahead manner in a coordinated manner of supply and demand.

Description

Technical Field

[0001] The present invention belongs to the technical field of smart heating, and specifically relates to a day-ahead optimization scheduling method for a heating system taking into account the coordination of supply and demand. Background Art

[0002] As centralized heating is one of the infrastructures of northern cities in my country, improving its intelligence level is an indispensable part of smart cities. Under the background of widely promoting the construction of smart cities, completing the intelligent upgrade of the heating system and realizing smart heating has also become an inevitable trend. Smart heating is a deep integration of heating technology and information technology. It is a modern integrated heating solution that integrates heat source production output, heating information collection, heating network monitoring, heating network hydraulic analysis, and room temperature collection. It achieves the goals of efficient, stable, low-carbon and energy-saving heating systems through real-time data collection, intelligent operation and control, and rational resource allocation.

[0003] In order to achieve the goal of smart heating, heating companies usually need to coordinate the supply and demand of heat from heating stations and heat sources one day in advance, formulate day-ahead scheduling plans, and solve the contradiction between the heat supply from heat sources and the heat demand from heating stations. This can play an important role in optimizing the scheduling of heat in the heating system. Advanced operation scheduling strategies are means that directly affect the heating goals and energy-saving effects of the heating system. However, in the actual heating process, due to the lack of a complete scheduling plan, heating management personnel usually roughly schedule the heat parameters of the heating station according to weather changes or rely on manual experience to adapt to changes in heating loads. This extensive scheduling method often results in heat not being distributed on demand, and the lack of supply and demand coordination between heat sources and heating stations, which reduces the scheduling efficiency and accuracy of the heating station and increases heating energy consumption.

[0004] Based on the above technical problems, it is necessary to design a new day-ahead optimal scheduling method for the heating system that considers the coordination of supply and demand. Summary of the invention

[0005] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a method for optimizing the day-ahead scheduling of a heating system taking into account the coordination of supply and demand, which can realize the coordination of supply and demand of heat source heat supply, heat demand of a thermal power station and actual heat supply of a thermal power station and the day-ahead scheduling of heat. There is no need for manual participation in the entire scheduling process, and the day-ahead scheduling plan is completely formulated by the machine independently. Manual operations are only performed for supervision, thereby saving heating energy and improving the efficiency and accuracy of scheduling.

[0006] In order to solve the above technical problems, the technical solution of the present invention is:

[0007] The present invention provides a method for optimizing the scheduling of a heating system based on supply and demand coordination, which comprises:

[0008] Step S1, determine the heating target and heating working conditions of each heating station, use machine learning algorithm to train the historical heating station load impact data, establish the load prediction model of each heating station for the next day, and obtain the substation demand load curve of each heating station for each time period of the next day;

[0009] Step S2, aligning and superimposing the demand loads of each substation in each period of the next day according to the substation demand load curves of each period of the next day, and establishing the total demand load curve of the heat source in each period of the next day;

[0010] Step S3, based on the heat source constraint conditions, the transmission delay of the heat network and the transmission heat loss factors, the total demand load curve of the heat source in each period of the next day is modulated to obtain the total supply load curve of the heat source in each period of the next day;

[0011] Step S4, automatically generating the substation supply load curves of each heating station in each time period of the next day according to the total supply load curves of the heat source in each time period of the next day and the substation demand load curves of each heating station in each time period of the next day, and after performing on-demand regulation of each heating station in each time period of the next day, generating the regulation instructions of each heating station in each time period of the next day;

[0012] Step S5: Based on the total supply load curve of the heat source in each time period the next day, the substation supply load curve of each heating station in each time period the next day, and the control instructions of each heating station in each time period the next day, a day-ahead optimization scheduling plan is constructed to achieve day-ahead optimization scheduling of the heating system with coordinated supply and demand.

[0013] Furthermore, the heating targets of each heating station include heating temperature, heating flow rate and heating unit consumption of the heating station;

[0014] The heating operating conditions of each thermal power station include meteorological information, time characteristic information and site characteristic information; the meteorological information includes outdoor temperature, humidity, wind speed, wind direction and light; the time characteristic information includes the early and late cold periods of heating, holidays, weekdays and weekends; the site characteristic information includes the site energy-saving characteristics, the site's hardware equipment attributes and the site's heating area.

[0015] Furthermore, the machine learning algorithm is used to train the historical thermal power station load impact data, establish the load prediction model of each thermal power station on the next day, and obtain the substation demand load curve of each thermal power station in each time period on the next day, including:

[0016] Obtain historical thermal power station load impact data, including the heating target, heating operating conditions, heat load of each thermal power station in each historical period, and the operating conditions of each period of the next day;

[0017] After data preprocessing and feature extraction of historical thermal power station load impact data, key data features affecting thermal power station load are obtained;

[0018] The key data characteristics of each historical period are used as the input variables of the load forecasting model of each thermal power station on the next day, and the heat load of each period on the next day is used as the output variable of the load forecasting model of each thermal power station on the next day;

[0019] Input variables are fed into the machine learning algorithm for model training to establish a load forecasting model for each thermal power station for the next day;

[0020] Based on the load prediction model of each thermal power station on the next day, the substation demand load change trend of each thermal power station in each time period of the next day is analyzed to obtain the substation demand load curve of each thermal power station in each time period of the next day; the time period of the next day is set according to the day-ahead optimization scheduling cycle.

[0021] Furthermore, the data preprocessing includes data missing value filling processing, outlier detection and correction processing, and normalization processing; the feature extraction is to use a CNN convolutional neural network to extract features from the preprocessed data, the activation function of the convolution layer 1 and the convolution layer 2 in the CNN convolutional neural network is Relu, the pooling layer selects the Maxpooling pooling method, and the fully connected layer flattens the data features from the convolution layer;

[0022] When the input variables are input into the machine learning algorithm for model training, a combined prediction method combining a GRU gated recurrent network and an election mechanism is adopted: the data features after feature extraction are used as input variables and input into the GRU gated recurrent network for model training to obtain the load forecast value of each heating station in each time period of the next day; a similar point selection algorithm is adopted to construct a similar heat load feature state set in historical data through a coarse clustering stage, and the category closest to the predicted heating station is selected as a new search space, and then through a similar point selection stage, the similarity between the heating station and the sample point in the coarse set is calculated, and the highest value is selected as the final similar point, and is recommended as the historical reference value; the weight parameter of the load forecast value of each heating station in each time period of the next day and the historical reference value in the election mechanism is determined; the load forecast value of each heating station in each time period of the next day and the historical reference value are input into the election mechanism, and the recommended value recommended by the two is used as the final load forecast value of each heating station in each time period of the next day.

[0023] Furthermore, after obtaining the substation demand load curves of each heating station in each time period of the next day, the method further includes using a KF Kalman filter algorithm to correct the substation demand load curves of each heating station in each time period of the next day, including:

[0024] Determine the state quantity in the KF Kalman filter algorithm, predict the state value X of the next period of the next day based on the state value of the current period of the next day, and write the state prediction equation and the prediction error covariance equation;

[0025] The load forecasting model of each thermal power station for the next day is established through machine learning, and the forecast value for the next period of the next day is obtained, which is used as the observation value Y. The state value X of the next period of the next day is used to update the corresponding error analysis matrix, and the optimal estimate of the next period of the next day is obtained at the same time.

[0026] The optimal substation demand load for each period of the next day is obtained by recursive calculation in sequence, and the substation demand load curve of each thermal power station in each period of the next day is corrected.

[0027] Further, the step S3 comprises:

[0028] During the operation of the heating system, the heat source operation constraints include the heat load variation range constraints and the heat source unit load increase and decrease rate constraints, which are expressed as:

[0029] D min ≤D≤D max ; D is the heat load of the heat source; D min and D max are the minimum and maximum heat loads that the heat source can provide, D max Depends on the design capacity of the heat source, D min It refers to the minimum heat load that a heat source can provide to ensure continuous, safe and stable operation;

[0030] ΔT is the day-ahead optimization scheduling cycle interval; δ D D is the maximum heat load rise and fall rate that the heat source can withstand; T The amount of heat that the heat source can provide during period T;

[0031] Establish the heat network transmission delay equation, including:

[0032] Calculate the hot water flow rate of the pipe network, expressed as:

[0033]

[0034]

[0035] v ps,k,t is the flow rate of hot water in the kth section of the water supply pipeline in time period t; ρ is the density of hot water; d k is the inner diameter of the kth section of the pipeline; q ps,k,t is the flow rate of the kth water supply pipeline in period t; λ is the unit conversion factor; v pr,k,t is the flow rate of hot water in the kth section of the return pipe during period t; q pr,k,t is the flow rate of the kth section of the return pipe in period t;

[0036] The hot water flow rate constraint is expressed as:

[0037]

[0038]

[0039] and are the lower and upper limits of the hot water flow rate of the kth water supply pipeline in time period t; and are the lower and upper limits of the hot water flow rate in the kth section of the return pipe during time period t;

[0040] Calculate the pipeline hot water transmission time, expressed as:

[0041]

[0042]

[0043] τ ps,k,t is the transmission time of the kth water supply pipeline in period t; j is the length of the jth section of the pipeline; S ps,k is the set of pipes between the hot water source and the kth section of the water supply pipe; τ pr,k,t is the transmission time of the kth section of the return pipe in period t; S pr,k It is the collection of pipes between the hot water source and the kth section of the return pipe;

[0044] The pipe hot water transmission time is rounded off and expressed as:

[0045]

[0046]

[0047] is the transmission time period of the kth water supply pipeline in period t; is the transmission time period of the kth section of the return pipe in period t;

[0048] After considering the transmission delay and heat loss of the heat network, the heat at the pipeline inlet and outlet meets the constraints, which can be expressed as:

[0049]

[0050]

[0051] is the heat at the inlet of water supply pipe k1 during period t; For water supply pipe k2 Heat at the outlet of the time period; μ hn is the heat loss rate of the heating network; is the heat at the inlet of the return pipe k1 during period t; For the return pipe k2 Heat at the time period outlet;

[0052] Based on the heat source constraint conditions, the transmission delay and transmission heat loss factors of the heat network, combined with the pipeline inlet and outlet heat constraint conditions, the modulation algorithm in the preset knowledge base is used to modulate the total demand load curve for each time period of the heat source the next day, eliminate the disturbance effect when the heat source is transmitted to the thermal power station, and obtain the total supply load curve for each time period of the heat source the next day; the modulation algorithm is based on a pre-established simulation model of heat source to heat network transmission of the heating system, sets the heat source constraint conditions, the transmission delay and transmission heat loss of the heat network, and sets the inlet and outlet heat constraint conditions of the heating pipeline, conducts multiple experiments and uses machine learning and parameter identification and correction algorithms to establish the relationship between the total demand load curve for each time period of the heat source the next day and the total supply load curve for each time period of the heat source the next day.

[0053] Further, the step S4 comprises:

[0054] According to the total supply load curve of the heat source in each period of the next day and the substation demand load curve of each heating station in each period of the next day, and with the control goal of achieving heat distribution on demand and supply-demand balance of each heating station, the operation constraints of the heating station are set, and the substation supply load curve of each heating station in each period of the next day is generated after solving the substation supply load of each heating station in each period of the next day using the preset control algorithm;

[0055] Based on the substation supply load curve of each thermal power station in each time period of the next day, the valve opening of each thermal power station is calculated to obtain the corresponding opening adjustment instructions, and the substation regulation in each time period of the next day is carried out, and the substation opening adjustment instructions in each time period of the next day are used as the regulation instructions of each thermal power station in each time period of the next day.

[0056] Furthermore, the control objective is to realize the heat distribution on demand and the balance of supply and demand of each thermal power station, set the thermal power station operation constraint conditions, and use the preset control algorithm to solve the substation supply load of each thermal power station in each period of the next day, including:

[0057] The control goal of achieving the on-demand distribution of heat at each heating station is expressed by calculating the relative error between the actual heat load of each heating station at each time period the next day and the substation demand load of each heating station at each time period the next day;

[0058] The control target of achieving heat supply and demand balance of each thermal power station is expressed by calculating the supply and demand balance between the total supply load of the heat source in each time period of the next day and the total demand load of all thermal power stations in each time period of the next day, and calculating the supply and demand balance between the substation demand load of each thermal power station in each time period of the next day and the substation supply load of each thermal power station in each time period of the next day; wherein, the total demand load of all thermal power stations in each time period of the next day is obtained by aligning, superimposing and summarizing the substation demand load curves of each thermal power station in each time period of the next day: first calculate the substation demand load of each thermal power station in each time period of the next day, and then superimpose and summarize the substation demand load of each thermal power station in each time period of the next day to obtain the total demand load of all thermal stations in each time period of the next day;

[0059] Determine the target weights of the heat demand allocation control target and the heat supply and demand balance control target of each thermal power station and aggregate them into a single target;

[0060] Set the operation constraints of the thermal power station, including the primary water supply temperature constraint of the thermal power station, the secondary return water temperature constraint of the thermal power station and the heat supply limit constraint of the thermal power station;

[0061] Based on the control objectives of heat distribution on demand and supply-demand balance of each thermal power station and the operating constraints of the thermal power station, the substation supply load model of each thermal power station in each period of the next day is established;

[0062] The preset control algorithm is used to solve the substation supply load model of each thermal power station in each time period the next day, and the substation supply load value of each thermal power station in each time period the next day is obtained.

[0063] Further, based on the substation supply load curves of each thermal power station in each period of the next day, the valve opening of each thermal power station is calculated to obtain the corresponding opening adjustment instruction, and the substation regulation in each period of the next day is performed, including:

[0064] Based on the substation supply load curves of each heating station in each period of the next day, the substation supply load values ​​of each heating station in each period of the next day are obtained, and a preset adjustment algorithm is used to establish an opening adjustment model between the substation supply load values ​​of each heating station in each period of the next day and the valve opening of each heating station;

[0065] Based on the opening regulation model, the valve opening regulation instructions of each thermal power station in each time period of the next day are obtained, and substation regulation in each time period is carried out the next day.

[0066] The present invention also provides a day-ahead optimal scheduling system for a heating system taking into account supply-demand coordination, which comprises:

[0067] The next day's substation demand load calculation unit is used to determine the heating target and heating operating conditions of each heating station, use machine learning algorithms to train historical heating station load impact data, establish the next day's load prediction model for each heating station, and obtain the substation demand load curves of each heating station in each period of the next day;

[0068] The next day's total demand load calculation unit for heat sources is used to align and superimpose the demand loads of each substation in each period of the next day according to the substation demand load curves of each period of the next day, and then establish the total demand load curves of the heat sources in each period of the next day;

[0069] The heat source load modulation unit is used to modulate the total demand load curve of the heat source in each period of the next day based on the heat source constraint conditions, the transmission delay of the heat network and the transmission heat loss factors, so as to obtain the total supply load curve of the heat source in each period of the next day;

[0070] The next day substation supply load calculation unit is used to automatically generate the substation supply load curves of each heating station in each period of the next day according to the total supply load curves of the heat source in each period of the next day and the substation demand load curves of each heating station in each period of the next day;

[0071] The substation control unit is used to control each heating station on demand in each period of the next day according to the substation supply load curve in each period of the next day, and generate control instructions for each heating station in each period of the next day;

[0072] The day-ahead dispatch plan generation unit is used to construct a day-ahead optimized dispatch plan based on the total supply load curve of the heat source in each time period the next day, the sub-station supply load curve of each heating station in each time period the next day, and the control instructions of each heating station in each time period the next day, so as to realize the day-ahead optimized dispatch of the heating system with coordinated supply and demand.

[0073] The beneficial effects of the present invention are:

[0074] The present invention determines the heating target and heating working conditions of each heating station, uses a machine learning algorithm to train the historical heating station load impact data, establishes a load prediction model for each heating station on the next day, and obtains the substation demand load curve of each heating station in each time period on the next day; after aligning and superimposing the demand load of each substation in each time period on the next day, the total demand load curve of the heat source in each time period on the next day is established; based on the heat source constraint conditions, the transmission delay of the heat network and the transmission heat loss factors, the total demand load curve of the heat source in each time period on the next day is calibrated. After modulation, the total supply load curve of the heat source for each period of the next day is obtained; according to the total supply load curve of the heat source for each period of the next day and the substation demand load curve of each heating station for each period of the next day, the substation supply load curve of each heating station for each period of the next day is automatically generated, and after the heating station is regulated on demand for each period of the next day, the regulation instructions for each heating station for each period of the next day are generated; according to the total supply load curve of the heat source for each period of the next day, the substation supply load curve of each heating station for each period of the next day and the regulation instructions for each heating station for each period of the next day, the day-ahead optimal dispatch forecast is constructed. The scheme can realize the optimal scheduling of the heating system with coordinated supply and demand. First, starting from the heat demand of the heating station one day in advance, the load prediction model of each heating station on the next day is established to obtain the sub-station demand load curve of each heating station in each time period on the next day. Then, based on the sub-station demand load curve in each time period on the next day, from the perspective of heat source heating and heat network heat transmission, the total supply load curve of the heat source in each time period on the next day is obtained, which can realize the coordinated supply and demand of the heat demand of the heating station and the heat supply of the heat source. Second, the total supply load curve of the heat source in each time period on the next day Based on the substation demand load curves of each thermal power station in each time period of the next day, the substation supply load curves of each thermal power station in each time period of the next day are generated, which can realize the supply and demand coordination of heat source heat supply, thermal power station heat demand and actual heat supply of thermal power station. A scheduling plan is formulated for the heating enterprise personnel one day in advance to realize the reasonable day-ahead optimization scheduling of system heat source and thermal power station heat. There is no need for human participation in the entire scheduling process. The day-ahead scheduling plan is formulated entirely by the machine independently, and humans only perform supervisory operations, saving heating energy and improving the efficiency and accuracy of scheduling.

[0075] Other features and advantages will be described in the following description, and partly become apparent from the description, or understood by practicing the invention. The purpose and other advantages of the invention are realized and obtained by the structures particularly pointed out in the description and the drawings.

[0076] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0078] Figure 1 This is a flow chart of a method for optimizing the day-ahead scheduling of a heating system considering the coordination of supply and demand according to the present invention;

[0079] Figure 2 This is a schematic diagram of the principle of day-ahead optimization scheduling of a heating system considering supply and demand coordination according to the present invention;

[0080] Figure 3 The figure is a schematic diagram of the structure of a day-ahead optimization scheduling system for a heating system taking into account the coordination of supply and demand according to the present invention. DETAILED DESCRIPTION

[0081] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0082] Example 1

[0083] Figure 1 It is a flow chart of a method for optimizing the day-ahead scheduling of a heating system taking into account the coordination of supply and demand involved in the present invention.

[0084] Figure 2 It is a schematic diagram of the day-ahead optimization scheduling principle of a heating system taking into account the coordination of supply and demand involved in the present invention.

[0085] like Figure 1-2 As shown, this embodiment 1 provides a method for optimizing the day-ahead scheduling of a heating system considering the coordination of supply and demand, which includes:

[0086] Step S1, determine the heating target and heating working conditions of each heating station, use machine learning algorithm to train the historical heating station load impact data, establish the load prediction model of each heating station for the next day, and obtain the substation demand load curve of each heating station for each time period of the next day;

[0087] Step S2, aligning and superimposing the demand loads of each substation in each period of the next day according to the substation demand load curves of each period of the next day, and establishing the total demand load curve of the heat source in each period of the next day;

[0088] Step S3, based on the heat source constraint conditions, the transmission delay of the heat network and the transmission heat loss factors, the total demand load curve of the heat source in each period of the next day is modulated to obtain the total supply load curve of the heat source in each period of the next day;

[0089] Step S4, automatically generating the substation supply load curves of each heating station in each time period of the next day according to the total supply load curves of the heat source in each time period of the next day and the substation demand load curves of each heating station in each time period of the next day, and after performing on-demand regulation of each heating station in each time period of the next day, generating the regulation instructions of each heating station in each time period of the next day;

[0090] Step S5: Based on the total supply load curve of the heat source in each time period the next day, the substation supply load curve of each heating station in each time period the next day, and the control instructions of each heating station in each time period the next day, a day-ahead optimization scheduling plan is constructed to achieve day-ahead optimization scheduling of the heating system with coordinated supply and demand.

[0091] It should be noted that step S2 first calculates the demand load of each substation during period t on the next day, and then superimposes and summarizes the demand load of each substation during period t on the next day to obtain the total demand load curve of the thermal power station during period t on the next day. Without considering the heat source constraint conditions, the transmission delay of the heating network and the transmission heat loss factors, the total demand load curve of the thermal power station during period t on the next day is equivalent to the total demand load curve of the heat source during period t on the next day; however, in the actual operation of the heating system, in order to achieve the accuracy of the day-ahead scheduling of the heating system thermal power station, it is necessary to adopt the heat source constraint conditions, the transmission delay of the heating network and the transmission loss factors in step S3. In actual applications, it is not limited to the heat source constraint conditions, the transmission delay of the heating network and the transmission loss factors, but may also include other factors, such as the heat medium temperature constraint of the pipeline, the pipeline flow constraint, etc.

[0092] In this embodiment, the heating targets of each heating station include heating temperature, heating flow rate and heating unit consumption of the heating station;

[0093] The heating operating conditions of each thermal power station include meteorological information, time characteristic information and site characteristic information; the meteorological information includes outdoor temperature, humidity, wind speed, wind direction and light; the time characteristic information includes the early and late cold periods of heating, holidays, weekdays and weekends; the site characteristic information includes the site energy-saving characteristics, the site's hardware equipment attributes and the site's heating area.

[0094] In this embodiment, the machine learning algorithm is used to train the historical thermal power station load impact data, establish the load prediction model of each thermal power station on the next day, and obtain the substation demand load curve of each thermal power station in each period of the next day, including:

[0095] Obtain historical thermal power station load impact data, including the heating target, heating operating conditions, heat load of each thermal power station in each historical period, and the operating conditions of each period of the next day;

[0096] After data preprocessing and feature extraction of historical thermal power station load impact data, key data features affecting thermal power station load are obtained;

[0097] The key data characteristics of each historical period are used as the input variables of the load forecasting model of each thermal power station on the next day, and the heat load of each period on the next day is used as the output variable of the load forecasting model of each thermal power station on the next day;

[0098] Input variables are fed into the machine learning algorithm for model training to establish a load forecasting model for each thermal power station for the next day;

[0099] Based on the load prediction model of each thermal power station on the next day, the substation demand load change trend of each thermal power station in each time period of the next day is analyzed to obtain the substation demand load curve of each thermal power station in each time period of the next day; the time period of the next day is set according to the day-ahead optimization scheduling cycle.

[0100] In this embodiment, the data preprocessing includes data missing value filling processing, outlier detection and correction processing, and normalization processing; the feature extraction is to use a CNN convolutional neural network to extract features from the preprocessed data, the activation function of the convolution layer 1 and the convolution layer 2 in the CNN convolutional neural network is Relu, the pooling layer selects the Maxpooling pooling method, and the fully connected layer flattens the data features from the convolution layer;

[0101] When the input variables are input into the machine learning algorithm for model training, a combined prediction method combining a GRU gated recurrent network and an election mechanism is adopted: the data features after feature extraction are used as input variables and input into the GRU gated recurrent network for model training to obtain the load forecast value of each heating station in each time period of the next day; a similar point selection algorithm is adopted to construct a similar heat load feature state set in historical data through a coarse clustering stage, and the category closest to the predicted heating station is selected as a new search space, and then through a similar point selection stage, the similarity between the heating station and the sample point in the coarse set is calculated, and the highest value is selected as the final similar point, and is recommended as the historical reference value; the weight parameter of the load forecast value of each heating station in each time period of the next day and the historical reference value in the election mechanism is determined; the load forecast value of each heating station in each time period of the next day and the historical reference value are input into the election mechanism, and the recommended value recommended by the two is used as the final load forecast value of each heating station in each time period of the next day.

[0102] It should be noted that the CNN convolutional neural network extracts feature vectors, which can be expressed as:

[0103]

[0104] p1=max(C1)+b2;

[0105]

[0106] p2=max(C2)+b4;

[0107] H c =f(p2×W3+b5)=Sigmoid(p2×W3+b5)=[h c1 ...h ct-1 ...h ct ...h ci ] T ;

[0108] X is the input variable; the output of convolution layer 1 is C1; the output of convolution layer 2 is C2; the output of pooling layer 1 is p1; the output of pooling layer 2 is p2; W1, W2, W3 are weight matrices; b1, b2, b3, b4, b5 are bias terms; max() is the maximum value function; H c is the feature vector after feature extraction.

[0109] The election mechanism is to establish a combined model of predicted values ​​and historical reference values. The historical reference value focuses on the similarity with the predicted thermal site, and selects the most similar sample point in the historical data as the historical reference value; the election function used to determine the final predicted recommended value is defined in the election mechanism, which is expressed as:

[0110] A=α*pre+β*his;

[0111] A is the final recommended value of the heat load prediction value and the heat load historical reference value of the fusion model; pre is the load prediction value of each heating station in each time period of the next day obtained by using the CNN-GRU model; his is the load historical reference value of each heating station in each time period of the next day; α and β are the respective proportions of the load prediction value and the historical reference value in the election mechanism.

[0112] The coarse clustering stage includes:

[0113] Establish the fuzzy matrix X of the historical load sample set of each thermal power station, where n is the total number of samples, q is the feature dimension, and x ij is the characteristic index j of sample point i;

[0114]

[0115] Initialize the relevant parameters of fuzzy clustering: number of clusters c, centroid matrix V, fuzzy index m, threshold ε, number of iterations L; the number of clusters c is calculated by [c min ,c max], with the minimum index XB as the goal, it is obtained by traversal search; V selects the initial center according to the principle of maximum difference probability pd; among them, the index XB comprehensively considers the similarity between samples of the same category and the difference between samples of different categories. The smaller the calculation result of XB, the better the clustering effect; when implementing clustering, the principle of making the distance between centroids as large as possible is adopted, and a method of selecting the point farthest from the existing center as the new cluster center each time is proposed. In order to measure the degree of distance, the distance proportion is defined To express the proportion of the distance between the sample point x and the center v to the total sample distance under the center, and then the sample for each center Perform product operation, and its value is used as the dissimilarity probability pd. It can express the degree of dissimilarity between the selected sample point and all the determined cluster centers, and is a comprehensive evaluation of the degree of distance;

[0116] Set the objective function, U is the membership matrix; u ij For sample x j Belongs to cluster v i The degree of membership; V is the centroid matrix, d ij is the Euclidean distance between samples;

[0117] The membership matrix U and the centroid matrix V are continuously updated until the convergence condition is reached or the maximum number of iterations is reached, and the final clustering results U and V are obtained;

[0118] According to the value in U, each sample corresponds to a sequence of centroid membership, and then the sample is divided into the cluster corresponding to the largest value in the sequence, so as to achieve sample classification;

[0119] Calculate the distance between the sample and each centroid, and select the nearest cluster as the new search space;

[0120] The similarity selection phase includes:

[0121] The samples in the coarse set are used as the sequence to be selected, and then the feature matching coefficient of the sequence to be selected and the target sequence and the time matching degree between the two are solved to obtain the similarity;

[0122] Compare the similarities of each sample, select the one with the largest similarity value as the similarity point, and take its real heat load value as the historical reference value;

[0123] The respective weights α and β of the load forecast value and the historical reference value in the election mechanism are determined: each sample is composed of the heat load forecast value, the historical reference value and the real load value. The intelligent optimization algorithm is used to obtain the parameter combination with the minimum objective function F(α,β), which is determined as the weight parameter of the election function;

[0124]

[0125] N is the number of samples, and Z is the current real load value. The fitness function F(α, β) is continuously iterated and calculated through the intelligent optimization algorithm, and the weight parameter combination corresponding to the minimum fitness value in the current iteration process is selected to update the parameters of the intelligent optimization algorithm. The search process is cyclically executed until the end condition is met, and the optimal solution of α and β is obtained, which is used as the weight parameter determined in the election mechanism.

[0126] In practical applications, based on the similarity principle of thermal station loads and the idea of ​​combined prediction, a combined prediction model based on CNN-GRU and election mechanism was established. The election mechanism was set as the core of the combination method, which was used to combine the model prediction results with the historical reference values ​​in the form of weight distribution. The obtained results were used as the final thermal station load values, which could combine the overall change law and the similarity of individual thermal stations, realize the joint effect of learning ability and historical experience, and improve the accuracy and stability of thermal load prediction.

[0127] In this embodiment, after obtaining the substation demand load curves of each heating station in each time period of the next day, the substation demand load curves of each heating station in each time period of the next day are corrected by using a KF Kalman filter algorithm, including:

[0128] Determine the state quantity in the KF Kalman filter algorithm, predict the state value X of the next period of the next day based on the state value of the current period of the next day, and write the state prediction equation and the prediction error covariance equation;

[0129] The load forecasting model of each thermal power station for the next day is established through machine learning, and the forecast value for the next period of the next day is obtained, and the forecast value is used as the observation value Y. The state value X of the next period of the next day is used to update the corresponding error analysis matrix, and the optimal estimate of the next period of the next day is obtained at the same time.

[0130] The optimal substation demand load for each period of the next day is obtained by recursive calculation in sequence, and the substation demand load curve of each thermal power station in each period of the next day is corrected.

[0131] It should be noted that in the actual operation of the heating system, when analyzing and processing the collected data, it is bound to be interfered by noise, including the measurement of noise, which requires real-time state update to reduce the impact of noise. Kalman filter KF is an optimal estimation method. By introducing the concepts of state variables and state space, the observation equation and state equation are established, and the unbiased minimum mean square error estimation is used to obtain the best prediction value after filtering the noise. In the measurement that completely contains noise, the state quantity estimated at the previous moment is matched with the observation quantity at this time to complete the update of the state variable, and at the same time, the estimated value at this time is obtained, which is convenient for real-time correction of load prediction. At the same time, the Kalman filter algorithm can make a good estimate of the dynamic system and eliminate the noise in the system.

[0132] The state equation and observation equation are expressed as:

[0133] X k+1 =AX k +w k ;

[0134] Y k =HX k +v k ;

[0135] X k is the system state, Y k is the system state X at time k k The observed value, A is the one-step transfer matrix of the system state, H is the observation matrix of the system, w k and v k are the random process noise and observation noise of the system respectively.

[0136] The Kalman filtering process includes:

[0137] Calculate the initial estimate for step k+1: X k+1|k =AX k ;

[0138] Update the covariance matrix: P k+1|k =AP k A T +Q;

[0139] Calculate the innovation: η k+1 =Y k+1|k -Y k =HX k+1|k -Y k ;

[0140] Calculate the Kalman gain matrix: K k+1 =P k+1|k H T (HP k+1|k H T +R) -1 ;

[0141] Calculate the optimal prediction value for k+1 steps: X k+1|k+1 =X k+1|k +K k+1 Y k+1 -K k+1 HX k+1|k ;

[0142] Update the covariance matrix for the next step: P k+1|k+1 =P k+1|k -K k HP k+1|k ;

[0143] Xk+1|k is the initial estimate for the k+1 step; X k+1|k+1 is the optimal prediction value for the next step; K k is the Kalman filter gain at time k; P k+1|k+1 is the covariance matrix for the next step; R and Q are two uncorrelated Gaussian noises w k and v k The covariance of .

[0144] In this embodiment, step S3 includes:

[0145] During the operation of the heating system, the heat source operation constraints include the heat load variation range constraints and the heat source unit load increase and decrease rate constraints, which are expressed as:

[0146] D min ≤D≤D max ; D is the heat load of the heat source; D min and D max are the minimum and maximum heat loads that the heat source can provide, D max Depends on the design capacity of the heat source, D min It refers to the minimum heat load that a heat source can provide to ensure continuous, safe and stable operation;

[0147] ΔT is the day-ahead optimization scheduling cycle interval; δ D D is the maximum heat load rise and fall rate that the heat source can withstand; T The amount of heat that the heat source can provide during period T;

[0148] Establish the heat network transmission delay equation, including:

[0149] Calculate the hot water flow rate of the pipe network, expressed as:

[0150]

[0151]

[0152] v ps,k,t is the flow rate of hot water in the kth section of the water supply pipeline in time period t; ρ is the density of hot water; d k is the inner diameter of the kth section of the pipeline; q ps,k,t is the flow rate of the kth water supply pipeline in period t; λ is the unit conversion factor; v pr,k,t is the flow rate of hot water in the kth section of the return pipe during period t; q pr,k,t is the flow rate of the kth section of the return pipe in period t;

[0153] The hot water flow rate constraint is expressed as:

[0154]

[0155]

[0156] and are the lower and upper limits of the hot water flow rate of the kth water supply pipeline in time period t; and are the lower and upper limits of the hot water flow rate in the kth section of the return pipe during time period t;

[0157] Calculate the pipeline hot water transmission time, expressed as:

[0158]

[0159]

[0160] τ ps,k,t is the transmission time of the kth water supply pipeline in period t; j is the length of the jth section of the pipeline; S ps,k is the set of pipes between the hot water source and the kth section of the water supply pipe; τ pr,k,t is the transmission time of the kth section of the return pipe in period t; S pr,k It is the collection of pipes between the hot water source and the kth section of the return pipe;

[0161] The pipe hot water transmission time is rounded off and expressed as:

[0162]

[0163]

[0164] is the transmission time period of the kth water supply pipeline in period t; is the transmission time period of the kth section of the return pipe in period t;

[0165] After considering the transmission delay and heat loss of the heat network, the heat at the pipeline inlet and outlet meets the constraints, which can be expressed as:

[0166]

[0167]

[0168] is the heat at the inlet of water supply pipe k1 during period t; For water supply pipe k2 Heat at the outlet of the time period; μ hn is the heat loss rate of the heating network; is the heat at the inlet of the return pipe k1 during period t; For the return pipe k2 Heat at the time period outlet;

[0169] Based on the heat source constraint conditions, the transmission delay and transmission heat loss factors of the heat network, combined with the pipeline inlet and outlet heat constraint conditions, the modulation algorithm in the preset knowledge base is used to modulate the total demand load curve for each time period of the heat source the next day, eliminate the disturbance effect when the heat source is transmitted to the thermal power station, and obtain the total supply load curve for each time period of the heat source the next day; the modulation algorithm is based on a pre-established simulation model of heat source to heat network transmission of the heating system, sets the heat source constraint conditions, the transmission delay and transmission heat loss of the heat network, and sets the inlet and outlet heat constraint conditions of the heating pipeline, conducts multiple experiments and uses machine learning and parameter identification and correction algorithms to establish the relationship between the total demand load curve for each time period of the heat source the next day and the total supply load curve for each time period of the heat source the next day.

[0170] In this embodiment, step S4 includes:

[0171] According to the total supply load curve of the heat source in each period of the next day and the substation demand load curve of each heating station in each period of the next day, and with the control goal of achieving heat distribution on demand and supply-demand balance of each heating station, the operation constraints of the heating station are set, and the substation supply load curve of each heating station in each period of the next day is generated after solving the substation supply load of each heating station in each period of the next day using the preset control algorithm;

[0172] Based on the substation supply load curve of each thermal power station in each time period of the next day, the valve opening of each thermal power station is calculated to obtain the corresponding opening adjustment instructions, and the substation regulation in each time period of the next day is carried out, and the substation opening adjustment instructions in each time period of the next day are used as the regulation instructions of each thermal power station in each time period of the next day.

[0173] In this embodiment, the control goal is to achieve heat distribution on demand and supply-demand balance in each thermal power station, set thermal power station operation constraints, and use a preset control algorithm to solve the substation supply load of each thermal power station in each period of the next day, including:

[0174] The control goal of achieving the on-demand distribution of heat at each heating station is expressed by calculating the relative error between the actual heat load of each heating station at each time period the next day and the substation demand load of each heating station at each time period the next day;

[0175] The control target of achieving heat supply and demand balance of each thermal power station is expressed by calculating the supply and demand balance between the total supply load of the heat source in each time period of the next day and the total demand load of all thermal power stations in each time period of the next day, and calculating the supply and demand balance between the substation demand load of each thermal power station in each time period of the next day and the substation supply load of each thermal power station in each time period of the next day; wherein, the total demand load of all thermal power stations in each time period of the next day is obtained by aligning, superimposing and summarizing the substation demand load curves of each thermal power station in each time period of the next day: first calculate the substation demand load of each thermal power station in each time period of the next day, and then superimpose and summarize the substation demand load of each thermal power station in each time period of the next day to obtain the total demand load of all thermal stations in each time period of the next day;

[0176] Determine the target weights of the heat demand allocation control target and the heat supply and demand balance control target of each thermal power station and aggregate them into a single target;

[0177] Set the operation constraints of the thermal power station, including the primary water supply temperature constraint of the thermal power station, the secondary return water temperature constraint of the thermal power station and the heat supply limit constraint of the thermal power station;

[0178] Based on the control objectives of heat distribution on demand and supply-demand balance of each thermal power station and the operating constraints of the thermal power station, the substation supply load model of each thermal power station in each period of the next day is established;

[0179] The preset control algorithm is used to solve the substation supply load model of each thermal power station in each time period the next day, and the substation supply load value of each thermal power station in each time period the next day is obtained.

[0180] It should be noted that the preset control algorithms for solving the substation supply load model of each thermal power station in each period of the next day include intelligent optimization algorithms, deep reinforcement learning algorithms and reinforcement learning algorithms. With the control goal of realizing the on-demand distribution of heat and supply-demand coordination of each thermal power station, taking into account the on-demand distribution of the thermal power station's own load, the supply and demand balance of the heat source and the thermal power station, and the supply and demand balance of the thermal power station's own substation supply and substation demand, the actual substation supply load values ​​of each thermal power station in each period of the next day are calculated, which can realize the reasonable scheduling of the system heat source and the thermal power station heat, save heating energy, and provide heating company personnel with scheduling plans in advance.

[0181] In this embodiment, based on the substation supply load curve of each thermal power station in each time period of the next day, the valve opening of each thermal power station is calculated to obtain the corresponding opening adjustment instruction, and the substation regulation of each time period of the next day is performed, including:

[0182] Based on the substation supply load curves of each heating station in each period of the next day, the substation supply load values ​​of each heating station in each period of the next day are obtained, and a preset adjustment algorithm is used to establish an opening adjustment model between the substation supply load values ​​of each heating station in each period of the next day and the valve opening of each heating station;

[0183] Based on the opening regulation model, the valve opening regulation instructions of each thermal power station in each time period of the next day are obtained, and substation regulation in each time period is carried out the next day.

[0184] It should be noted that the opening regulation model is established based on the historical data of valve opening of the thermal power station, the historical load of the thermal power station, the operating conditions and other conditions, using machine learning algorithms, deep learning algorithms and model prediction algorithms.

[0185] Example 2

[0186] Figure 3 It is a structural schematic diagram of a day-ahead optimization scheduling system for a heating system taking into account the coordination of supply and demand involved in the present invention.

[0187] like Figure 3 As shown, this embodiment 2 provides a day-ahead optimal scheduling system for a heating system considering supply and demand coordination, which includes:

[0188] The next day's substation demand load calculation unit is used to determine the heating target and heating operating conditions of each heating station, use machine learning algorithms to train historical heating station load impact data, establish the next day's load prediction model for each heating station, and obtain the substation demand load curves of each heating station in each period of the next day;

[0189] The next day's total demand load calculation unit for heat sources is used to align and superimpose the demand loads of each substation in each period of the next day according to the substation demand load curves of each period of the next day, and then establish the total demand load curves of the heat sources in each period of the next day;

[0190] The heat source load modulation unit is used to modulate the total demand load curve of the heat source in each period of the next day based on the heat source constraint conditions, the transmission delay of the heat network and the transmission heat loss factors, so as to obtain the total supply load curve of the heat source in each period of the next day;

[0191] The next day substation supply load calculation unit is used to automatically generate the substation supply load curves of each heating station in each period of the next day according to the total supply load curves of the heat source in each period of the next day and the substation demand load curves of each heating station in each period of the next day;

[0192] The substation control unit is used to control each heating station on demand in each period of the next day according to the substation supply load curve in each period of the next day, and generate control instructions for each heating station in each period of the next day;

[0193] The day-ahead dispatch plan generation unit is used to construct a day-ahead optimized dispatch plan based on the total supply load curve of the heat source in each time period the next day, the sub-station supply load curve of each heating station in each time period the next day, and the control instructions of each heating station in each time period the next day, so as to realize the day-ahead optimized dispatch of the heating system with coordinated supply and demand.

[0194] In several embodiments provided in the present application, it should be understood that the disclosed system and method can also be implemented in other ways. The system embodiments described above are merely schematic. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the systems, methods and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and a part of the module, program segment or code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0195] In addition, the functional modules in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0196] If the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0197] Based on the above ideal embodiments of the present invention, the relevant staff can make various changes and modifications without departing from the technical concept of the present invention through the above description. The technical scope of the present invention is not limited to the contents of the specification, and its technical scope must be determined according to the scope of the claims.

Claims

1. A day-ahead optimization scheduling method for a heating system considering supply and demand coordination, characterized in that: It includes: Step S1, determine the heating target and heating working conditions of each heating station, use machine learning algorithm to train the historical heating station load impact data, establish the load prediction model of each heating station for the next day, and obtain the substation demand load curve of each heating station for each time period of the next day; Step S2, aligning and superimposing the demand loads of each substation in each period of the next day according to the substation demand load curves of each period of the next day, and establishing the total demand load curve of the heat source in each period of the next day; Step S3, based on the heat source constraint conditions, the transmission delay of the heat network and the transmission heat loss factors, the total demand load curve of the heat source in each period of the next day is modulated to obtain the total supply load curve of the heat source in each period of the next day; Step S4, automatically generating the substation supply load curves of each heating station in each time period of the next day according to the total supply load curves of the heat source in each time period of the next day and the substation demand load curves of each heating station in each time period of the next day, and after performing on-demand regulation of each heating station in each time period of the next day, generating the regulation instructions of each heating station in each time period of the next day; Step S5: Based on the total supply load curve of the heat source in each time period the next day, the substation supply load curve of each heating station in each time period the next day, and the control instructions of each heating station in each time period the next day, a day-ahead optimization scheduling plan is constructed to achieve day-ahead optimization scheduling of the heating system with coordinated supply and demand.

2. The method for optimizing the day-ahead scheduling of a heating system according to claim 1, characterized in that: The heating targets of each heating station include heating temperature, heating flow rate and heating unit consumption of the heating station; The heating operating conditions of each thermal power station include meteorological information, time characteristic information and site characteristic information; the meteorological information includes outdoor temperature, humidity, wind speed, wind direction and light; the time characteristic information includes the early and late cold periods of heating, holidays, weekdays and weekends; the site characteristic information includes the site energy-saving characteristics, the site's hardware equipment attributes and the site's heating area.

3. The method for optimizing the day-ahead scheduling of a heating system according to claim 1, characterized in that: The machine learning algorithm is used to train the historical thermal power station load impact data, establish the load prediction model of each thermal power station on the next day, and obtain the substation demand load curve of each thermal power station in each period of the next day, including: Obtain historical thermal power station load impact data, including the heating target, heating operating conditions, heat load of each thermal power station in each historical period, and the operating conditions of each period of the next day; After data preprocessing and feature extraction of historical thermal power station load impact data, key data features affecting thermal power station load are obtained; The key data characteristics of each historical period are used as the input variables of the load forecasting model of each thermal power station on the next day, and the heat load of each period on the next day is used as the output variable of the load forecasting model of each thermal power station on the next day; Input variables are fed into the machine learning algorithm for model training to establish a load forecasting model for each thermal power station for the next day; Based on the load prediction model of each thermal power station on the next day, the substation demand load change trend of each thermal power station in each time period of the next day is analyzed to obtain the substation demand load curve of each thermal power station in each time period of the next day; the time period of the next day is set according to the day-ahead optimization scheduling cycle.

4. The method for optimizing the day-ahead scheduling of a heating system according to claim 3, characterized in that: The data preprocessing includes data missing value filling, outlier detection and correction, and normalization. The feature extraction is to use a CNN convolutional neural network to extract features from the preprocessed data. The activation functions of the convolutional layer 1 and the convolutional layer 2 in the CNN convolutional neural network are Relu, the pooling layer selects the Maxpooling pooling method, and the fully connected layer flattens the data features from the convolutional layer. When the input variables are input into the machine learning algorithm for model training, a combined prediction method combining a GRU gated recurrent network and an election mechanism is adopted: the data features after feature extraction are used as input variables and input into the GRU gated recurrent network for model training to obtain the load forecast value of each heating station in each period of the next day; a similar point selection algorithm is adopted to construct a similar heat load feature state set in historical data through a coarse clustering stage, and the category closest to the predicted heating station is selected as a new search space, and then through a similar point selection stage, the similarity between the heating station and the sample point in the coarse set is calculated, and the highest value is selected as the final similarity point, and is recommended as a historical reference value; Determine the weight parameters of the load forecast value and historical reference value of each heating station in each period of the next day in the election mechanism; The load forecast value and historical reference value of each heating station for each time period of the next day are input into the election mechanism, and the recommended value selected by the two is used as the final load forecast value of each heating station for each time period of the next day.

5. The method for optimizing the day-ahead scheduling of a heating system according to claim 3, characterized in that: After obtaining the substation demand load curves of each heating station in each time period of the next day, the method further includes using a KF Kalman filter algorithm to correct the substation demand load curves of each heating station in each time period of the next day, including: Determine the state quantity in the KF Kalman filter algorithm, predict the state value X of the next period of the next day based on the state value of the current period of the next day, and write the state prediction equation and the prediction error covariance equation; The load forecasting model of each thermal power station for the next day is established through machine learning, and the forecast value for the next period of the next day is obtained, and the forecast value is used as the observation value Y. The state value X of the next period of the next day is used to update the corresponding error analysis matrix, and the optimal estimate of the next period of the next day is obtained at the same time. The optimal substation demand load for each period of the next day is obtained by recursive calculation in sequence, and the substation demand load curve of each thermal power station in each period of the next day is corrected.

6. The method for optimizing the day-ahead scheduling of a heating system according to claim 1, characterized in that: The step S3 comprises: During the operation of the heating system, the heat source operation constraints include the heat load variation range constraints and the heat source unit load increase and decrease rate constraints, which are expressed as: D min ≤D≤D max ; D is the heat load of the heat source; D min and D max are the minimum and maximum heat loads that the heat source can provide, D max Depends on the design capacity of the heat source, D min It refers to the minimum heat load that a heat source can provide to ensure continuous, safe and stable operation; ΔT is the day-ahead optimization scheduling cycle interval; δ D D is the maximum heat load rise and fall rate that the heat source can withstand; T The amount of heat that the heat source can provide during period T; Establish the heat network transmission delay equation, including: Calculate the hot water flow rate of the pipe network, expressed as: v ps,k,t is the flow rate of hot water in the kth section of the water supply pipeline in time period t; ρ is the density of hot water; d k is the inner diameter of the kth section of the pipeline; q ps,k,t is the flow rate of the kth water supply pipeline in period t; λ is the unit conversion factor; v pr,k,t is the flow rate of hot water in the kth section of the return pipe during period t; q pr,k,t is the flow rate of the kth section return pipe in period t; The hot water flow rate constraint is expressed as: and are the lower and upper limits of the hot water flow rate of the kth water supply pipeline in time period t; and are the lower and upper limits of the hot water flow rate in the kth section of the return pipe during time period t; Calculate the pipeline hot water transmission time, expressed as: τ ps,k,t is the transmission time of the kth water supply pipeline in period t; j is the length of the jth section of the pipeline; S ps,k is the set of pipes between the hot water source and the kth section of the water supply pipe; τ pr,k,t is the transmission time of the kth section of the return pipe in period t; S pr,k It is the collection of pipes between the hot water source and the kth section of the return pipe; The pipe hot water transmission time is rounded off and expressed as: is the transmission time period of the kth water supply pipeline in period t; is the transmission time period of the kth section of the return pipe in period t; After considering the transmission delay and heat loss of the heat network, the heat at the pipeline inlet and outlet meets the constraints, which can be expressed as: is the heat at the inlet of water supply pipe k1 during period t; For water supply pipe k2 Heat at the outlet of the time period; μ hn is the heat loss rate of the heating network; is the heat at the inlet of the return pipe k1 during period t; For the return pipe k2 Heat at the time period outlet; Based on the heat source constraint conditions, the transmission delay and transmission heat loss factors of the heat network, combined with the pipeline inlet and outlet heat constraint conditions, the modulation algorithm in the preset knowledge base is used to modulate the total demand load curve for each time period of the heat source the next day, eliminate the disturbance effect when the heat source is transmitted to the thermal power station, and obtain the total supply load curve for each time period of the heat source the next day; the modulation algorithm is based on a pre-established simulation model of heat source to heat network transmission of the heating system, sets the heat source constraint conditions, the transmission delay and transmission heat loss of the heat network, and sets the inlet and outlet heat constraint conditions of the heating pipeline, conducts multiple experiments and uses machine learning and parameter identification and correction algorithms to establish the relationship between the total demand load curve for each time period of the heat source the next day and the total supply load curve for each time period of the heat source the next day.

7. The method for optimizing the day-ahead scheduling of a heating system according to claim 1, characterized in that: The step S4 comprises: According to the total supply load curve of the heat source in each period of the next day and the substation demand load curve of each heating station in each period of the next day, and with the control goal of achieving heat distribution on demand and supply-demand balance of each heating station, the operation constraints of the heating station are set, and the substation supply load curve of each heating station in each period of the next day is generated after solving the substation supply load of each heating station in each period of the next day using the preset control algorithm; Based on the substation supply load curve of each thermal power station in each time period of the next day, the valve opening of each thermal power station is calculated to obtain the corresponding opening adjustment instructions, and the substation regulation in each time period of the next day is carried out, and the substation opening adjustment instructions in each time period of the next day are used as the regulation instructions of each thermal power station in each time period of the next day.

8. The method for optimizing the day-ahead scheduling of a heating system according to claim 7, characterized in that: The control goal is to achieve heat distribution on demand and supply-demand balance in each thermal power station, set the thermal power station operation constraints, and use a preset control algorithm to solve the substation supply load of each thermal power station in each period of the next day, including: The control goal of achieving the on-demand distribution of heat at each heating station is expressed by calculating the relative error between the actual heat load of each heating station at each time period the next day and the substation demand load of each heating station at each time period the next day; The control target of achieving heat supply and demand balance of each thermal power station is expressed by calculating the supply and demand balance between the total supply load of the heat source in each time period of the next day and the total demand load of all thermal power stations in each time period of the next day, and calculating the supply and demand balance between the substation demand load of each thermal power station in each time period of the next day and the substation supply load of each thermal power station in each time period of the next day; wherein, the total demand load of all thermal power stations in each time period of the next day is obtained by aligning, superimposing and summarizing the substation demand load curves of each thermal power station in each time period of the next day: first calculate the substation demand load of each thermal power station in each time period of the next day, and then superimpose and summarize the substation demand load of each thermal power station in each time period of the next day to obtain the total demand load of all thermal stations in each time period of the next day; Determine the target weights of the heat demand allocation control target and the heat supply and demand balance control target of each thermal power station and aggregate them into a single target; Set the operation constraints of the thermal power station, including the primary water supply temperature constraint of the thermal power station, the secondary return water temperature constraint of the thermal power station and the heat supply limit constraint of the thermal power station; Based on the control objectives of heat distribution on demand and supply-demand balance of each thermal power station and the operating constraints of the thermal power station, the substation supply load model of each thermal power station in each period of the next day is established; The preset control algorithm is used to solve the substation supply load model of each thermal power station in each time period the next day, and the substation supply load value of each thermal power station in each time period the next day is obtained.

9. The method for optimizing the day-ahead scheduling of a heating system according to claim 7, characterized in that: The method of calculating the valve opening of each thermal power station based on the substation supply load curve of each thermal power station in each time period of the next day to obtain the corresponding opening adjustment instruction and performing substation control in each time period of the next day includes: Based on the substation supply load curves of each heating station in each period of the next day, the substation supply load values ​​of each heating station in each period of the next day are obtained, and a preset adjustment algorithm is used to establish an opening adjustment model between the substation supply load values ​​of each heating station in each period of the next day and the valve opening of each heating station; Based on the opening regulation model, the valve opening regulation instructions of each thermal power station in each time period of the next day are obtained, and substation regulation in each time period is carried out the next day.

10. A day-ahead optimization scheduling system for a heating system considering supply and demand coordination, characterized in that: It includes: The next day's substation demand load calculation unit is used to determine the heating target and heating operating conditions of each heating station, use machine learning algorithms to train historical heating station load impact data, establish the next day's load prediction model for each heating station, and obtain the substation demand load curves of each heating station in each period of the next day; The next day's total demand load calculation unit for heat sources is used to align and superimpose the demand loads of each substation in each period of the next day according to the substation demand load curves of each period of the next day, and then establish the total demand load curves of the heat sources in each period of the next day; The heat source load modulation unit is used to modulate the total demand load curve of the heat source in each period of the next day based on the heat source constraint conditions, the transmission delay of the heat network and the transmission heat loss factors, so as to obtain the total supply load curve of the heat source in each period of the next day; The next day substation supply load calculation unit is used to automatically generate the substation supply load curves of each heating station in each period of the next day according to the total supply load curves of the heat source in each period of the next day and the substation demand load curves of each heating station in each period of the next day; The substation control unit is used to control each heating station on demand in each period of the next day according to the substation supply load curve in each period of the next day, and generate control instructions for each heating station in each period of the next day; The day-ahead dispatch plan generation unit is used to construct a day-ahead optimized dispatch plan based on the total supply load curve of the heat source in each time period the next day, the sub-station supply load curve of each heating station in each time period the next day, and the control instructions of each heating station in each time period the next day, so as to realize the day-ahead optimized dispatch of the heating system with coordinated supply and demand.

Images