Heat supply system secondary network regulation and control effect evaluation method based on hybrid algorithm

Abstract

The invention discloses a heat supply system secondary network regulation and control effect evaluation method based on a hybrid algorithm. The method comprises the following steps: constructing a regulation and control effect evaluation system; performing weight distribution on each evaluation index in the regulation and control effect evaluation system to obtain weight information of each evaluation index, and performing regulation and control effect evaluation grade division according to the weight information of each evaluation index and the selected secondary network regulation and control effect characteristic parameters; establishing a heat supply system secondary network regulation and control effect real-time evaluation model, and obtaining a first evaluation result of the regulation and control effect; and comprehensively evaluating the regulation and control effect of the secondary network from the aspect of a plurality of classified and set first evaluation indexes by adopting a fuzzy comprehensive evaluation method in combination with the weight information of each evaluation index to obtain a second evaluation result of the regulation and control effect, and according to the first evaluation result and the second evaluation result of the regulation and control effect, performing comprehensive evaluation of the regulation and control effect, judging whether the regulation and control effect accords with an expected regulation and control effect, and performing optimization adjustment of a secondary network regulation and control strategy to realize optimization of the secondary network regulation and control effect.

Description

Technical Field

[0001] The present invention belongs to the technical field of smart heating systems, and specifically relates to a method for evaluating the control effect of a secondary network of a heating system based on a hybrid algorithm. Background Art

[0002] Smart heating is a modern heating method based on digital, networked and intelligent information technology, with users as its target, low carbon, comfort and efficiency as its main features, and thorough perception, extensive interconnection and deep intelligence as its technical characteristics.

[0003] The secondary network control of the heating system usually uses indoor temperature sensing, temperature control equipment, cloud platform intelligent software, the Internet of Things and intelligent modeling technology based on heat network big data to achieve coordinated optimization and control of heat sources, thermal power stations, primary networks and secondary networks. It is an important component and key technology of smart heating, and pays more attention to the control of unit buildings and heat users on the secondary network side. Through intelligent algorithms, the opening of the electric control valve in front of the unit building and the opening of the electric control valve in front of the household are set to reasonably distribute the flow and heat of each branch line to meet the personalized heating needs of the unit building and heat users. However, at present, the evaluation of the effect of the secondary network after control is only a simple analysis through monitoring the heating operation data, or through the investigation of user satisfaction. There is no complete comprehensive evaluation system for the control effect of the secondary network, and it is impossible to evaluate the control effect in multiple dimensions.

[0004] Based on the above technical problems, it is necessary to design a new evaluation method for the secondary network control effect of the heating system based on a hybrid algorithm. 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 hybrid algorithm for evaluating the control effect of the secondary network of a heating system, which can establish a control effect evaluation system, conduct a comprehensive effect evaluation from multiple dimensions of various evaluation indicators, and ensure the accuracy of the evaluation results of the control effect of the secondary network, as well as optimize the control strategy based on the control effect evaluation results. On the one hand, it ensures the optimization of the control effect of the secondary network, and on the other hand, it can also obtain the optimal secondary network control strategy.

[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 evaluating the control effect of a secondary network of a heating system based on a hybrid algorithm, which comprises:

[0008] S1. Determine the evaluation object, basic attribute parameters, control operation mode and control operation strategy for the secondary network control of the heating system;

[0009] S2. Obtain the control operation data of the secondary network of the heating system when executing relevant control operation strategies in different evaluation object dimensions and different control operation modes, screen out the evaluation features with great correlation with the control effect, and classify and set the first-level evaluation indicators of the control effect, decompose the first-level evaluation indicators downward, set the corresponding second-level evaluation indicators, and build a control effect evaluation system;

[0010] S3. Use the indicator weight determination method to assign weights to each evaluation indicator in the regulation effect evaluation system, obtain the weight information of each evaluation indicator, and divide the regulation effect evaluation level according to the weight information of each evaluation indicator and the selected secondary network regulation effect characteristic parameters;

[0011] S4. Based on the deep learning algorithm, combined with the weight information of each evaluation index, the changes in the characteristic parameters of the secondary network control effect during the monitoring and control process, and the divided control effect evaluation levels, a real-time evaluation model for the control effect of the secondary network of the heating system is established after training and learning, and the first evaluation result of the control effect is obtained;

[0012] And, using a fuzzy comprehensive evaluation method combined with the weight information of each evaluation index, comprehensively evaluating the secondary network regulation effect from the aspects of multiple first evaluation indicators set by classification to obtain a second evaluation result of the regulation effect;

[0013] S5. After conducting a comprehensive evaluation of the regulation and control effects based on the first and second evaluation results, determine whether it meets the expected regulation and control effects, and optimize and adjust the secondary network regulation and control strategies to achieve the optimization of the secondary network regulation and control effects.

[0014] Further, in S1, the evaluation objects include heat users, unit buildings, communities and heat stations;

[0015] The control operation mode is set according to different evaluation objects, including household control operation mode, building control operation mode, community control operation mode and heating station control operation mode;

[0016] The control operation strategy is set based on the control operation mode and related basic attribute parameters, as well as external environmental parameters and historical operation data; the basic attribute parameters include the indoor heating area of ​​the heat user, the location and orientation of the heat user's room, the heat usage behavior habits of the heat user, the heating area of ​​the unit building, the location and orientation of the unit building, the overall heating area of ​​the community, the location and orientation of the community and the scale and layout of the thermal power station.

[0017] Further, in S2, the control operation data at least includes heat source heat generation, energy consumption, carbon dioxide emissions, heating network temperature, pressure, flow, user indoor temperature, user satisfaction survey, complaint rate and hydraulic balance data, and the time for control to reach stability;

[0018] The screening out of evaluation features with a high correlation with the regulation effect and the classification setting of the first-level evaluation indicators of the regulation effect include: using the correlation analysis method to calculate the correlation coefficient between each regulation operation data and the secondary network regulation effect, and screening out the evaluation features with a high correlation with the secondary network regulation effect; the correlation between each regulation operation data and the secondary network regulation effect is related to the multi-dimensional factors of the evaluation object, the regulation operation mode, and the regulation operation strategy, and the evaluation features of the regulation effect are different for different evaluation objects, regulation operation modes and regulation operation strategies; and calculating the similarity between each evaluation feature, presetting the number of clusters and using the clustering algorithm to perform multiple clustering, to obtain the classification setting of the first-level evaluation indicators of the regulation effect.

[0019] Furthermore, the first-level evaluation indicators include control technology indicators, control energy-saving indicators, control economic indicators, control environment indicators and control satisfaction indicators;

[0020] The second-level evaluation indicators corresponding to the control technology indicators include: control stability, control hydraulic imbalance, control time, deviation between actual control parameters and expected target parameters, and control automation equipment coverage rate;

[0021] The second-level evaluation indicators corresponding to the energy-saving control indicators include: energy utilization rate, energy consumption per unit area, heat consumption, water consumption and electricity consumption;

[0022] The second-level evaluation indicators corresponding to the control economic indicators include: fixed costs, operating costs and maintenance costs;

[0023] The second-level evaluation indicators corresponding to the regulatory environment indicators include: reduction ratio of carbon dioxide emissions and emissions of harmful substances;

[0024] The second-level evaluation indicators corresponding to the regulation satisfaction indicators include: user satisfaction score and user complaint rate.

[0025] Furthermore, in S3, the weight of each evaluation indicator in the control effect evaluation system is assigned using an indicator weight determination method to obtain weight information of each evaluation indicator, including:

[0026] The AHP method is used to calculate the subjective weight of each evaluation index in the second-level evaluation index;

[0027] The critic method is used to calculate the objective weight of each evaluation index in the second-level evaluation index;

[0028] The game theory method is used to comprehensively calculate the subjective weight and objective weight of each evaluation indicator, realize the weight distribution of each evaluation indicator, and obtain the weight calculation of each evaluation indicator.

[0029] Furthermore, in S3, the regulation effect evaluation level is divided according to the weight information of each evaluation indicator and the selected secondary network regulation effect characteristic parameters, including: presetting the scoring criteria for each evaluation indicator according to expert experience, including the maximum value, minimum value and scoring mechanism of each evaluation indicator; performing weighted summation of the scores of each evaluation indicator according to the scoring criteria of each evaluation indicator and combining the weight information of each evaluation indicator to obtain the regulation effect evaluation score, and dividing the evaluation into different levels according to the evaluation score.

[0030] Furthermore, the different levels based on the evaluation scores include: Level I excellent, Level II excellent, Level III good, Level IV qualified and Level V unqualified, and the levels decrease from high to low.

[0031] Furthermore, in S4, a real-time evaluation model for the regulation effect of the secondary network of the heating system is established, including:

[0032] The evaluation data including the weight information of each evaluation indicator, the changes in the characteristic parameters of the secondary network regulation effect during the monitoring and control process, and the evaluation level of the regulation effect are used as sample data of the evaluation model. The training set and the test set are divided. The training set is input into the BiLSTM model for training and learning, and then a real-time evaluation model for the regulation effect of the secondary network of the heating system is established. The test set is used to verify the evaluation performance of the real-time evaluation model for the regulation effect of the secondary network and adjust the model parameters.

[0033] Furthermore, in S4, a fuzzy comprehensive evaluation method is used in combination with the weight information of each evaluation index to comprehensively evaluate the secondary network regulation effect from the aspects of multiple first evaluation indicators set by classification, including:

[0034] The fuzzy relationship matrix is ​​constructed using the membership function and is expressed as:

[0035]

[0036] r ij is the affiliation between the i-th evaluation indicator and the j-th regulation effect evaluation level;

[0037] According to the fuzzy relationship matrix and the weight information of each evaluation index, a fuzzy evaluation vector B = A·R is established; A is the weight vector synthesized by the weights of each evaluation index;

[0038] The comprehensive evaluation of the secondary network regulation effect is based on the fuzzy evaluation vector, which is expressed as:

[0039]

[0040] n is the number of levels; b is the number of levels; n is the fuzzy evaluation vector belonging to the nth evaluation level; k is the undetermined coefficient; x nis the evaluation score of the nth level; M is the fuzzy comprehensive evaluation result of the secondary network regulation effect.

[0041] Furthermore, in said S5, the methods used for comprehensive evaluation of the regulation effect include weighted average method, hierarchical analysis method, entropy weight method, and machine learning algorithm;

[0042] The determination of whether it meets the expected effect of regulation and control, and optimizing and adjusting the secondary network regulation and control strategy to achieve the optimization of the secondary network regulation and control effect include: judging the indicators of the first-level evaluation indicators and the second-level evaluation indicators that do not meet the expected effect of regulation and control based on the comprehensive evaluation results of the regulation and control effect and the scores of each evaluation indicator, and optimizing and adjusting the secondary network regulation and control strategy based on the indicator, executing the regulation and control strategy for a preset time period, continuing to execute S4 to evaluate the regulation and control effect and executing S5 to determine whether it meets the expected effect of regulation and control, repeating the execution until it meets the expected effect of regulation and control, and achieving the optimization of the secondary network regulation and control effect.

[0043] The beneficial effects of the present invention are:

[0044] (1) The present invention can set corresponding control operation modes and control operation strategies according to different evaluation objects by determining the evaluation objects, basic attribute parameters, control operation modes and control operation strategies during the control of the secondary network of the heating system. Because different objects are under different heating stages or under the influence of other external conditions, the control operation strategies of the secondary network are adapted and dynamically changed;

[0045] (2) The present invention obtains the control operation data of the secondary network of the heating system when executing relevant control operation strategies under different evaluation object dimensions and different control operation modes, screens out the evaluation features with great correlation with the control effect, and classifies and sets the first-level evaluation indicators of the control effect, and decomposes the first-level evaluation indicators downward, sets the corresponding second-level evaluation indicators, and constructs a control effect evaluation system; it can screen out the evaluation features with great correlation with the control effect based on the collected and obtained relevant control operation data, reduce the evaluation feature dimensions, and facilitate the subsequent comprehensive control effect evaluation from different dimensions and different evaluation indicators by classifying and setting the first-level evaluation indicators and setting the corresponding second-level evaluation indicators;

[0046] (3) The present invention uses an indicator weight determination method to assign weights to each evaluation indicator in the regulation effect evaluation system, obtains weight information of each evaluation indicator, and divides the regulation effect evaluation level according to the weight information of each evaluation indicator and the selected secondary network regulation effect characteristic parameters; it can ensure that the evaluation system more accurately reflects the actual situation of the regulation effect, enhances the objectivity and fairness of the evaluation, and clarifies which evaluation indicators are more important, so as to facilitate more attention in the subsequent evaluation and regulation strategy optimization;

[0047] (4) The present invention establishes a real-time evaluation model for the regulation effect of the secondary network of the heating system after training and learning based on a deep learning algorithm combined with the weight information of each evaluation index, the changes in the characteristic parameters of the regulation effect of the secondary network during the monitoring and control process, and the divided regulation effect evaluation levels, and obtains a first evaluation result of the regulation effect; and, adopts a fuzzy comprehensive evaluation method combined with the weight information of each evaluation index to comprehensively evaluate the regulation effect of the secondary network from the aspects of multiple first evaluation indicators set by classification to obtain a second evaluation result of the regulation effect; the deep learning algorithm can be used to realize the initial real-time evaluation of the regulation effect of the secondary network, and the changes in the regulation process can be quickly responded to, and an immediate evaluation result can be provided to adapt to the constantly changing characteristics of the regulation effect, and a secondary evaluation is performed through the fuzzy comprehensive evaluation method, and the effect evaluation is performed from various dimensions, which can also be regarded as an offline evaluation result;

[0048] (5) The present invention conducts a comprehensive evaluation of the regulation and control effect based on the first evaluation result and the second evaluation result of the regulation and control effect, determines whether it meets the expected regulation and control effect, and optimizes and adjusts the secondary network regulation and control strategy to achieve the optimization of the secondary network regulation and control effect; it can comprehensively evaluate the two evaluation results to ensure the accuracy of the evaluation results of the secondary network regulation and control effect, and optimize the regulation and control strategy according to the evaluation results of the regulation and control effect. On the one hand, it ensures the optimization of the secondary network regulation and control effect, and on the other hand, it can also obtain the optimal secondary network regulation and control strategy.

[0049] 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.

[0050] 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

[0051] 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.

[0052] Figure 1 This is a flow chart of a method for evaluating the control effect of a secondary network of a heating system based on a hybrid algorithm according to the present invention;

[0053] Figure 2 Schematic diagram of the control effect evaluation system of the present invention. DETAILED DESCRIPTION

[0054] 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.

[0055] Example 1

[0056] like Figure 1 , Figure 2 As shown, this embodiment 1 provides a method for evaluating the control effect of a secondary network of a heating system based on a hybrid algorithm, which includes:

[0057] S1. Determine the evaluation object, basic attribute parameters, control operation mode and control operation strategy for the secondary network control of the heating system;

[0058] S2. Obtain the control operation data of the secondary network of the heating system when executing relevant control operation strategies in different evaluation object dimensions and different control operation modes, screen out the evaluation features with great correlation with the control effect, and classify and set the first-level evaluation indicators of the control effect, decompose the first-level evaluation indicators downward, set the corresponding second-level evaluation indicators, and build a control effect evaluation system;

[0059] S3. Use the indicator weight determination method to assign weights to each evaluation indicator in the regulation effect evaluation system, obtain the weight information of each evaluation indicator, and divide the regulation effect evaluation level according to the weight information of each evaluation indicator and the selected secondary network regulation effect characteristic parameters;

[0060] S4. Based on the deep learning algorithm, combined with the weight information of each evaluation index, the changes in the characteristic parameters of the secondary network control effect during the monitoring and control process, and the divided control effect evaluation levels, a real-time evaluation model for the control effect of the secondary network of the heating system is established after training and learning, and the first evaluation result of the control effect is obtained;

[0061] And, using a fuzzy comprehensive evaluation method combined with the weight information of each evaluation index, comprehensively evaluating the secondary network regulation effect from the aspects of multiple first evaluation indicators set by classification to obtain a second evaluation result of the regulation effect;

[0062] S5. After conducting a comprehensive evaluation of the regulation and control effects based on the first and second evaluation results, determine whether it meets the expected regulation and control effects, and optimize and adjust the secondary network regulation and control strategies to achieve the optimization of the secondary network regulation and control effects.

[0063] In this embodiment, in S1, the evaluation objects include heat users, unit buildings, communities and heat stations;

[0064] The control operation mode is set according to different evaluation objects, including household control operation mode, building control operation mode, community control operation mode and heating station control operation mode;

[0065] The control operation strategy is set based on the control operation mode and related basic attribute parameters, as well as external environmental parameters and historical operation data; the basic attribute parameters include the indoor heating area of ​​the heat user, the location and orientation of the heat user's room, the heat usage behavior habits of the heat user, the heating area of ​​the unit building, the location and orientation of the unit building, the overall heating area of ​​the community, the location and orientation of the community and the scale and layout of the thermal power station.

[0066] In this embodiment, in S2, the control operation data at least includes heat source heat generation, energy consumption, carbon dioxide emissions, heating network temperature, pressure, flow, user indoor temperature, user satisfaction survey, complaint rate and hydraulic balance data, and the time for control to reach stability;

[0067] The screening out of evaluation features with a high correlation with the regulation effect and the classification setting of the first-level evaluation indicators of the regulation effect include: using the correlation analysis method to calculate the correlation coefficient between each regulation operation data and the secondary network regulation effect, and screening out the evaluation features with a high correlation with the secondary network regulation effect; the correlation between each regulation operation data and the secondary network regulation effect is related to the multi-dimensional factors of the evaluation object, the regulation operation mode, and the regulation operation strategy, and the evaluation features of the regulation effect are different for different evaluation objects, regulation operation modes and regulation operation strategies; and calculating the similarity between each evaluation feature, presetting the number of clusters and using the clustering algorithm to perform multiple clustering, to obtain the classification setting of the first-level evaluation indicators of the regulation effect.

[0068] It should be noted that the correlation analysis methods include Pearson correlation coefficient, Spearman rank correlation coefficient, Kendall rank correlation coefficient, and Pearson partial correlation coefficient.

[0069] The similarity between the evaluation features is calculated, the number of clusters is preset, and the clustering algorithm is used to perform multiple clusterings to obtain the first-level evaluation index classification setting of the regulation effect, including:

[0070] 1) Calculate similarity:

[0071] Similarity measurement: Select an appropriate similarity measurement method to calculate the similarity between features. Commonly used similarity measurement methods include Euclidean distance, cosine similarity, Jaccard similarity, etc.

[0072] Similarity matrix: construct a similarity matrix, each element in the matrix represents the similarity between two features;

[0073] 2) Preset number of clusters:

[0074] Determine the number of clusters: Determine the expected number of clusters based on domain knowledge or through some heuristic methods (such as elbow rule, silhouette coefficient, etc.);

[0075] 3) Select a clustering algorithm

[0076] Clustering algorithm: Select a suitable clustering algorithm. Common clustering algorithms include K-means, hierarchical clustering, DBSCAN, etc.

[0077] Algorithm selection: If the number of clusters has been preset, you can use the K-means algorithm; if the number of clusters is not sure, you can consider using hierarchical clustering or DBSCAN algorithm.

[0078] 4) Perform multiple clustering

[0079] Run multiple times: Since the results of some clustering algorithms (such as K-means) may be affected by initialization, it is recommended to run the clustering algorithm multiple times and select the best result;

[0080] Evaluate clustering quality: Use clustering quality evaluation indicators (such as silhouette coefficient, Calinski-Harabasz index, etc.) to evaluate the results of each clustering;

[0081] 5) Obtain the first-level evaluation index classification

[0082] Determine the optimal clustering result: select the optimal clustering result based on the evaluation index;

[0083] Classification setting: Map the clustering results to the first-level evaluation indicator classification setting.

[0084] In this embodiment, the first-level evaluation indicators include control technology indicators, control energy-saving indicators, control economic indicators, control environment indicators and control satisfaction indicators;

[0085] The second-level evaluation indicators corresponding to the control technology indicators include: control stability, control hydraulic imbalance, control time, deviation between actual control parameters and expected target parameters, and control automation equipment coverage rate;

[0086] The second-level evaluation indicators corresponding to the energy-saving control indicators include: energy utilization rate, energy consumption per unit area, heat consumption, water consumption and electricity consumption;

[0087] The second-level evaluation indicators corresponding to the control economic indicators include: fixed costs, operating costs and maintenance costs;

[0088] The second-level evaluation indicators corresponding to the regulatory environment indicators include: reduction ratio of carbon dioxide emissions and emissions of harmful substances;

[0089] The second-level evaluation indicators corresponding to the regulation satisfaction indicators include: user satisfaction score and user complaint rate.

[0090] In this embodiment, in S3, a weight allocation method is used to allocate weights to each evaluation indicator in the control effect evaluation system to obtain weight information of each evaluation indicator, including:

[0091] The AHP method is used to calculate the subjective weight of each evaluation index in the second-level evaluation index;

[0092] The critic method is used to calculate the objective weight of each evaluation index in the second-level evaluation index;

[0093] The game theory method is used to comprehensively calculate the subjective weight and objective weight of each evaluation indicator, realize the weight distribution of each evaluation indicator, and obtain the weight calculation of each evaluation indicator.

[0094] It should be noted that the AHP method is used to calculate the subjective weight of each evaluation index, including:

[0095] 1) Constructing the judgment matrix: By comparing the evaluation indicators in the control effect evaluation system in pairs, determine the relative importance, and then determine the values ​​of the relative importance of each evaluation indicator in pairs according to the nine-scale method, construct the judgment matrix B = (b ij ) n×n ; b ij is the importance of indicator i to indicator j, and n is the number of indicators;

[0096] 2) Calculate subjective weight: Normalize the judgment matrix and express it as:

[0097]

[0098] The subjective weight is expressed as:

[0099]

[0100] 3) Check consistency: The consistency index is expressed as:

[0101] C=(λ max -n) / (n-1);

[0102] λ max To determine the maximum eigenvalue of the matrix, when C approaches 0, the matrix shows good consistency. The larger C is, the less consistent the matrix is.

[0103] The critic method is used to calculate the objective weight of each evaluation indicator, including:

[0104] 1) The evaluation index is dimensionless and expressed as:

[0105] x ij ′=(x ij -x jmin ) / (x jmax -x jmin );

[0106] x ij ′=(x jmax -x ij ( / -x jmax -x jmin );

[0107] x ij ′ is the normalized indicator matrix element; x ij is the original matrix element variable value, the original data matrix is ​​X=(x ij ) m×n , m is the number of samples to be evaluated, n is the number of evaluation indicators; x jmin 、x jmax are the minimum and maximum values ​​of the sample of index j respectively;

[0108] 2) Calculate the index difference, conflict and information content, which are expressed as:

[0109]

[0110] S j is the dissimilarity of index j; x 0j is the mean value of each indicator;

[0111]

[0112] E j is the conflict of indicator j; r ij is the correlation coefficient between index i and j;

[0113] V j =S j E j ;

[0114] V j is the information content of indicator j;

[0115] 3) Calculate the objective weight, expressed as:

[0116]

[0117] W j is the objective weight of indicator j.

[0118] The game theory method is used to comprehensively calculate the subjective weight and objective weight of each evaluation index: the game theory idea is adopted, and the subjective weight and objective weight are used as the two sides of the game, so that the difference between the comprehensive weight and the subjective weight and the objective weight is minimized, and the combination coefficient is determined to form a comprehensive weight that complements the subjective and objective advantages;

[0119] Construct a weight vector set and use subjective and objective weighting methods for n indicators. Assume that the subjective weight is w 1 , the objective weight is w 2 , then the two sets of vectors are linearly combined a 1 and a 2 are the combination coefficients of subjective weight and objective weight respectively;

[0120] With weight vector W and w 1 、w 2 The deviation and minimization of the objective function is the coefficient a 1 and a 2 Optimize, expressed as:

[0121]

[0122] The optimal combination weight vector W obtained by solution analysis is expressed as:

[0123]

[0124] In this embodiment, in S3, the regulation effect evaluation level is divided according to the weight information of each evaluation indicator and the selected secondary network regulation effect characteristic parameters, including: presetting the scoring criteria for each evaluation indicator according to expert experience, including the maximum value, minimum value and scoring mechanism of each evaluation indicator; according to the scoring criteria for each evaluation indicator, combining the weight information of each evaluation indicator, weighted summing up the score of each evaluation indicator to obtain the regulation effect evaluation score, and dividing it into different levels according to the evaluation score.

[0125] In this embodiment, the different levels according to the evaluation scores include: Level I Excellent, Level II Excellent, Level III Good, Level IV Qualified and Level V Unqualified, and the levels decrease from high to low.

[0126] In this embodiment, in S4, a real-time evaluation model for the regulation effect of the secondary network of the heating system is established, including:

[0127] The evaluation data including the weight information of each evaluation indicator, the changes in the characteristic parameters of the secondary network regulation effect during the monitoring and control process, and the evaluation level of the regulation effect are used as sample data of the evaluation model. The training set and the test set are divided. The training set is input into the BiLSTM model for training and learning, and then a real-time evaluation model for the regulation effect of the secondary network of the heating system is established. The test set is used to verify the evaluation performance of the real-time evaluation model for the regulation effect of the secondary network and adjust the model parameters.

[0128] It should be noted that the hierarchical structure of the BiLSTM model includes:

[0129] BiLSTM Layer:

[0130] Receive training set data and use the two-layer BiLSTM to fully capture the context in the data. Use the intelligent optimization algorithm to optimize the hyperparameter combination of the BiLSTM layer to improve the prediction accuracy.

[0131] Dropout layer:

[0132] Randomly inactivate a certain proportion of neuron outputs to improve the generalization performance of the model and prevent overfitting;

[0133] Attention layer:

[0134] Attention is introduced to assign weights to the time-step hidden vector output by the Dropout layer, and the weighted summed hidden vector is output;

[0135] Dense layer:

[0136] The fully connected layer transforms the output dimension and obtains the evaluation result.

[0137] In this embodiment, in S4, a fuzzy comprehensive evaluation method is used in combination with the weight information of each evaluation index to comprehensively evaluate the secondary network control effect from the aspects of multiple first evaluation indicators set by classification, including:

[0138] The fuzzy relationship matrix is ​​constructed using the membership function and is expressed as:

[0139]

[0140] r ij is the affiliation between the i-th evaluation indicator and the j-th regulation effect evaluation level;

[0141] According to the fuzzy relationship matrix and the weight information of each evaluation index, a fuzzy evaluation vector B = A·R is established; A is the weight vector synthesized by the weights of each evaluation index;

[0142] The comprehensive evaluation of the secondary network regulation effect is based on the fuzzy evaluation vector, which is expressed as:

[0143]

[0144] n is the number of levels; b is the number of levels; n is the fuzzy evaluation vector belonging to the nth evaluation level; k is the undetermined coefficient; x n is the evaluation score of the nth level; M is the fuzzy comprehensive evaluation result of the secondary network regulation effect.

[0145] In this embodiment, in S5, the method used for comprehensive evaluation of the control effect includes weighted average method, hierarchical analysis method, entropy weight method, and machine learning algorithm;

[0146] The determination of whether it meets the expected effect of regulation and control, and optimizing and adjusting the secondary network regulation and control strategy to achieve the optimization of the secondary network regulation and control effect include: judging the indicators of the first-level evaluation indicators and the second-level evaluation indicators that do not meet the expected effect of regulation and control based on the comprehensive evaluation results of the regulation and control effect and the scores of each evaluation indicator, and optimizing and adjusting the secondary network regulation and control strategy based on the indicator, executing the regulation and control strategy for a preset time period, continuing to execute S4 to evaluate the regulation and control effect and executing S5 to determine whether it meets the expected effect of regulation and control, repeating the execution until it meets the expected effect of regulation and control, and achieving the optimization of the secondary network regulation and control effect.

[0147] 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.

[0148] In addition, the functional modules in each embodiment of the present invention can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part. If the function is implemented in the form of a software functional 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, in essence, 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, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a 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, RandomAccess Memory), disk or optical disk and other media that can store program codes.

[0149] 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 method for evaluating the control effect of a secondary network of a heating system based on a hybrid algorithm, characterized in that: It includes: S1. Determine the evaluation object, basic attribute parameters, control operation mode and control operation strategy for the secondary network control of the heating system; S2. Obtain the control operation data of the secondary network of the heating system when executing relevant control operation strategies in different evaluation object dimensions and different control operation modes, screen out the evaluation features with great correlation with the control effect, and classify and set the first-level evaluation indicators of the control effect, decompose the first-level evaluation indicators downward, set the corresponding second-level evaluation indicators, and build a control effect evaluation system; S3. Use the indicator weight determination method to assign weights to each evaluation indicator in the regulation effect evaluation system, obtain the weight information of each evaluation indicator, and divide the regulation effect evaluation level according to the weight information of each evaluation indicator and the selected secondary network regulation effect characteristic parameters; S4. Based on the deep learning algorithm, combined with the weight information of each evaluation index, the changes in the characteristic parameters of the secondary network control effect during the monitoring and control process, and the divided control effect evaluation levels, a real-time evaluation model for the control effect of the secondary network of the heating system is established after training and learning, and the first evaluation result of the control effect is obtained; And, using a fuzzy comprehensive evaluation method combined with the weight information of each evaluation index, comprehensively evaluating the secondary network regulation effect from the aspects of multiple first evaluation indicators set by classification to obtain a second evaluation result of the regulation effect; S5. After conducting a comprehensive evaluation of the regulation and control effects based on the first and second evaluation results, determine whether it meets the expected regulation and control effects, and optimize and adjust the secondary network regulation and control strategies to achieve the optimization of the secondary network regulation and control effects.

2. The method for evaluating the control effect of the secondary network of the heating system according to claim 1 is characterized in that: In S1, the evaluation objects include heat users, unit buildings, communities and heat stations; The control operation mode is set according to different evaluation objects, including household control operation mode, building control operation mode, community control operation mode and heating station control operation mode; The control operation strategy is set based on the control operation mode and related basic attribute parameters, as well as external environmental parameters and historical operation data; the basic attribute parameters include the indoor heating area of ​​the heat user, the location and orientation of the heat user's room, the heat usage behavior habits of the heat user, the heating area of ​​the unit building, the location and orientation of the unit building, the overall heating area of ​​the community, the location and orientation of the community and the scale and layout of the thermal power station.

3. The method for evaluating the control effect of the secondary network of the heating system according to claim 1 is characterized in that: In S2, the control operation data at least includes heat source heat generation, energy consumption, carbon dioxide emissions, heating network temperature, pressure, flow, user indoor temperature, user satisfaction survey, complaint rate and hydraulic balance data, and the time for control to reach stability; The screening out of the evaluation features with a high correlation with the regulation effect and the classification and setting of the first-level evaluation indicators of the regulation effect include: using the correlation analysis method to calculate the correlation coefficient between each regulation operation data and the secondary network regulation effect, and screening out the evaluation features with a high correlation with the secondary network regulation effect; The correlation between each control operation data and the control effect of the secondary network is related to the multi-dimensional factors of the evaluation object, control operation mode, and control operation strategy. Different evaluation objects, control operation modes and control operation strategies have different evaluation characteristics of the control effect; and the similarity between each evaluation feature is calculated, the number of clusters is preset, and the clustering algorithm is used for multiple clustering to obtain the first-level evaluation indicator classification setting of the control effect.

4. The method for evaluating the control effect of the secondary network of a heating system according to claim 1 is characterized in that: The first-level evaluation indicators include control technology indicators, control energy-saving indicators, control economic indicators, control environment indicators and control satisfaction indicators; The second-level evaluation indicators corresponding to the control technology indicators include: control stability, control hydraulic imbalance, control time, deviation between actual control parameters and expected target parameters, and control automation equipment coverage rate; The second-level evaluation indicators corresponding to the energy-saving control indicators include: energy utilization rate, energy consumption per unit area, heat consumption, water consumption and electricity consumption; The second-level evaluation indicators corresponding to the control economic indicators include: fixed costs, operating costs and maintenance costs; The second-level evaluation indicators corresponding to the regulatory environment indicators include: reduction ratio of carbon dioxide emissions and emissions of harmful substances; The second-level evaluation indicators corresponding to the regulation satisfaction indicators include: user satisfaction score and user complaint rate.

5. The method for evaluating the control effect of the secondary network of the heating system according to claim 1 is characterized in that: In S3, the weight of each evaluation indicator in the control effect evaluation system is assigned using an indicator weight determination method to obtain the weight information of each evaluation indicator, including: The AHP method is used to calculate the subjective weight of each evaluation index in the second-level evaluation index; The critic method is used to calculate the objective weight of each evaluation index in the second-level evaluation index; The game theory method is used to comprehensively calculate the subjective weight and objective weight of each evaluation indicator, realize the weight distribution of each evaluation indicator, and obtain the weight calculation of each evaluation indicator.

6. The method for evaluating the control effect of the secondary network of a heating system according to claim 1, characterized in that: In S3, the regulation effect evaluation level is divided according to the weight information of each evaluation indicator and the selected secondary network regulation effect characteristic parameters, including: presetting the scoring criteria of each evaluation indicator according to expert experience, including the maximum value, minimum value and scoring mechanism of each evaluation indicator; according to the scoring criteria of each evaluation indicator, combining the weight information of each evaluation indicator, weighted summing up the score of each evaluation indicator to obtain the regulation effect evaluation score, and dividing into different levels according to the evaluation score.

7. The method for evaluating the control effect of the secondary network of the heating system according to claim 6 is characterized in that: The evaluation scores are divided into different levels, including: Level I Excellent, Level II Excellent, Level III Good, Level IV Qualified and Level V Unqualified, and the levels decrease from high to low.

8. The method for evaluating the control effect of the secondary network of a heating system according to claim 1, characterized in that: In S4, a real-time evaluation model for the regulation effect of the secondary network of the heating system is established, including: The evaluation data including the weight information of each evaluation indicator, the changes in the characteristic parameters of the secondary network regulation effect during the monitoring and control process, and the evaluation level of the regulation effect are used as sample data of the evaluation model. The training set and the test set are divided. The training set is input into the BiLSTM model for training and learning, and then a real-time evaluation model for the regulation effect of the secondary network of the heating system is established. The test set is used to verify the evaluation performance of the real-time evaluation model for the regulation effect of the secondary network and adjust the model parameters.

9. The method for evaluating the control effect of the secondary network of a heating system according to claim 1, characterized in that: In S4, a fuzzy comprehensive evaluation method is used in combination with the weight information of each evaluation index to comprehensively evaluate the secondary network regulation effect from the aspects of multiple first evaluation indicators set by classification, including: The fuzzy relationship matrix is ​​constructed using the membership function and is expressed as: r ij is the affiliation between the i-th evaluation indicator and the j-th regulation effect evaluation level; According to the fuzzy relationship matrix and the weight information of each evaluation index, a fuzzy evaluation vector B = A·R is established; A is the weight vector synthesized by the weights of each evaluation index; The comprehensive evaluation of the secondary network regulation effect is based on the fuzzy evaluation vector, which is expressed as: n is the number of levels; b is the number of levels; n is the fuzzy evaluation vector belonging to the nth evaluation level; k is the undetermined coefficient; x n is the evaluation score of the nth level; M is the fuzzy comprehensive evaluation result of the secondary network regulation effect.

10. The method for evaluating the control effect of the secondary network of a heating system according to claim 1, characterized in that: In said S5, the methods used for comprehensive evaluation of the regulation effect include weighted average method, hierarchical analysis method, entropy weight method, and machine learning algorithm; The determination of whether it meets the expected effect of regulation and control, and optimizing and adjusting the secondary network regulation and control strategy to achieve the optimization of the secondary network regulation and control effect include: judging the indicators of the first-level evaluation indicators and the second-level evaluation indicators that do not meet the expected effect of regulation and control based on the comprehensive evaluation results of the regulation and control effect and the scores of each evaluation indicator, and optimizing and adjusting the secondary network regulation and control strategy based on the indicator, executing the regulation and control strategy for a preset time period, continuing to execute S4 to evaluate the regulation and control effect and executing S5 to determine whether it meets the expected effect of regulation and control, repeating the execution until it meets the expected effect of regulation and control, and achieving the optimization of the secondary network regulation and control effect.

Images