System architecture evaluation and optimization method based on structure entropy and centrality algorithm

Abstract

The invention provides a system architecture evaluation and optimization method based on a structure entropy and a centrality algorithm, and the method comprises the steps: firstly constructing different information transmission networks, and then employing the structure entropy algorithm to recognize an important path in any information transmission network and a non-important path needing to be simplified; a centrality algorithm is used to identify components with high comprehensive centrality and simplified components with low comprehensive centrality in any information transmission network, and the architecture of each information transmission network is optimized based on the identification result. Meanwhile, the complexity of different information transfer networks is measured by combining the order degree calculated by the structure entropy algorithm and the comprehensive centrality index calculated by the centrality algorithm, so that the different information transfer networks are evaluated. By means of the method, the information transmission efficiency and stability of the system can be improved, and therefore an efficient electric vehicle integrated heat management system optimization framework is established.

Description

Technical Field

[0001] The present invention relates to the technical field of integrated thermal management systems for electric vehicles, and in particular to a system architecture evaluation and optimization method based on structural entropy and centrality algorithms. Background Art

[0002] In the research field of integrated thermal management systems for electric vehicles, with the development of technology and the increase in market demand, the requirements for system performance are also increasing. The thermal management system must not only meet the needs of passenger comfort, but also ensure the thermal safety and performance stability of the battery, motor and electronic control system. Therefore, the architectural design of the thermal management system becomes particularly important.

[0003] Many studies have attempted to integrate multiple heat sources into one system to improve energy efficiency. The complexity of integrated thermal management systems increases with the number and type of heat sources, making system design and optimization more difficult. These integrated systems may adopt more complex loop design and control strategies, but often lack system-level performance evaluation and optimization methods. Existing evaluation methods often focus on thermodynamic performance, while ignoring the information transfer efficiency and overall complexity of the system architecture. Therefore, extending traditional thermodynamic performance analysis to the information domain can explain the system-level architecture optimization brought about by the development of integrated thermal management systems, and provide important practical significance for the research of new integrated thermal management systems in the future. Summary of the invention

[0004] In order to solve the deficiencies in the prior art, the present invention provides a system architecture evaluation and optimization method based on structural entropy and centrality algorithms. The present invention first constructs different information transmission networks, and then uses the structural entropy algorithm to identify the important paths and non-important paths that need to be simplified in any information transmission network; the centrality algorithm is used to identify the components with high comprehensive centrality and simplified components with low comprehensive centrality in any information transmission network, and the architecture of each information transmission network is optimized based on the identification results. At the same time, the order calculated by the structural entropy algorithm and the comprehensive centrality index calculated by the centrality algorithm are combined to measure the complexity of different information transmission networks to evaluate different information transmission networks. The above method can improve the information transmission efficiency and stability of the system, thereby establishing an efficient optimization architecture for the integrated thermal management system of electric vehicles.

[0005] In order to achieve the above object, the specific scheme adopted by the present invention is: A system architecture evaluation and optimization method based on structural entropy and centrality algorithm is used to evaluate and optimize the integrated thermal management system architecture of electric vehicles, which mainly includes the following steps: S1. Establish a model for the energy information coupling and transfer between multi-temperature heat sources during the fully coupled operation of the passenger compartment, power battery and motor electronic control thermal management requirements of pure electric vehicles; S2. Abstract the components in the thermal management system into network nodes, and then connect the network nodes based on different actual situations to form different information transmission networks, construct an undirected information structure graph based on the information transmission network, construct a directed information structure graph based on the information transmission network and the energy information coupling and transmission model established in step S1, and describe the information transmission relationship between the components based on the undirected information structure graph and the directed information structure graph; S3. Use the structural entropy algorithm to evaluate the total order of different information transmission networks to quantify the transmission efficiency and accuracy of each information transmission network; for different information transmission networks, according to the timeliness entropy and quality entropy of the two vertices in the node pair, obtain the efficiency and accuracy of different paths, use efficiency and accuracy as path importance evaluation indicators, and identify important paths and non-important paths that need to be simplified in any information transmission network; S4. Use the centrality algorithm to calculate the comprehensive centrality of each component in each information transmission network, use the comprehensive centrality of each component as an evaluation indicator of the role played by each component, determine the role of each component in information transmission in different information transmission networks, and identify the components with high comprehensive centrality and the components with low comprehensive centrality that need to be simplified in any information transmission network; S5. Combine the order calculated by the structural entropy algorithm and the comprehensive centrality index calculated by the centrality algorithm to measure the complexity of different information transmission networks, so as to evaluate different information transmission networks; S6. Based on the judgment results of step S3 and step S4, simplify non-important paths and components with low comprehensive centrality, and strengthen the connection between components with high comprehensive centrality to optimize the architecture of each information transmission network and improve the efficiency and stability of information transmission.

[0006] Furthermore, the centrality algorithm includes degree centrality, closeness centrality, and betweenness centrality.

[0007] Furthermore, the degree centrality, closeness centrality, and betweenness centrality are used to calculate the standard deviation of degree centrality, the standard deviation of closeness centrality, and the standard deviation of betweenness centrality, respectively. The three standard deviation values ​​are then normalized. After that, the normalized standard deviation of degree centrality, closeness centrality, and betweenness centrality can be used to calculate the comprehensive centrality σ. According to the calculation results of the comprehensive centrality, all components are ranked, and the components with the highest centrality indicators are identified as the key components.

[0008] Furthermore, the calculation formula of comprehensive centrality σ is: σ=ω D ·σ′ D +ω C ·σ′ C +ω B ·σ′B In the formula, σ′ D represents the normalized standard deviation of degree centrality, σ c ′ represents the normalized standard deviation of closeness centrality, σ′ B represents the normalized standard deviation of betweenness centrality, ω D ,ω C ,ω B are the weight coefficients of normalized degree centrality, normalized closeness centrality, and normalized betweenness centrality, respectively. D =0.3,ω C =0.4,ω B =0.3.

[0009] Furthermore, the order degree calculated by the structural entropy and the comprehensive centrality index calculated by the centrality algorithm are combined to calculate the complexity SC of the information transmission network: SC=γ(1-R)+ησ Where SC is the system complexity, R is the total order of the system, σ is the comprehensive centrality, γ and η represent the weight coefficients of the total order of the system and the comprehensive centrality, respectively, and both are taken as 0.5.

[0010] Beneficial effects:

[0011] (1) The system architecture evaluation and optimization method based on structural entropy and centrality algorithm proposed in the present invention is a novel method that not only takes into account the thermodynamic performance, but also comprehensively evaluates the information transmission efficiency and complexity of the system architecture.

[0012] (2) This method abstracts the components and paths in the thermal management system by constructing an information transmission network, providing a new perspective to analyze and optimize the system architecture.

[0013] (3) The application of structural entropy enables this method to quantify the order of different information transmission networks and identify bottlenecks and redundancies in information transmission, thereby providing a theoretical basis for the simplification and optimization of paths in different information transmission networks.

[0014] (4) Centrality algorithms can help designers identify key components, optimize connections among key components, and improve the overall performance and stability of the system by evaluating the importance of components in the information transmission network. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 Flowchart for structural entropy algorithm evaluation.

[0016] Figure 2 Evaluation flow chart of the centrality algorithm.

[0017] Figure 3It is a flow chart of the system architecture evaluation and optimization method proposed in the present invention. DETAILED DESCRIPTION

[0018] The technical solution of the present invention will be clearly and completely described below in conjunction with specific embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0019] This invention provides a system architecture evaluation and optimization based on structural entropy and centrality algorithm, please refer to Figure 3 , this method is based on structural entropy and centrality algorithms, focusing on optimizing the complex architecture of the thermal management system of pure electric vehicles. First, the energy information coupling and transmission model between multi-temperature heat sources during the fully coupled operation of the fully coupled thermal management requirements of the passenger compartment, power battery and motor electronic control of pure electric vehicles is constructed. Further, an information transmission network is constructed, the thermal management system components are abstracted as network nodes, and the information transmission relationship between components is displayed through undirected information structure diagrams and directed information structure diagrams. The application of the structural entropy method quantifies the total order and important paths of the system, while the centrality algorithm identifies components with high comprehensive centrality (i.e., key components), revealing the key control points and bottlenecks within the system. Based on these analyses, the present invention proposes an optimization strategy, which improves the information transmission efficiency and stability of the system by identifying and simplifying inefficient paths and non-important components, and strengthening the connection between key components when building the actual system architecture based on the analysis results. The method of the present invention provides designers with powerful tools to help identify and optimize key components and paths that affect system performance. The complexity of different architectures is identified, and an evaluation index for different system architecture models is proposed. The introduction of this novel methodology provides an important theoretical basis and practical guidance for the design and optimization of electric vehicle thermal management systems, and demonstrates a new direction for optimizing system architecture from the perspective of information flow. The specific steps are described in detail below.

[0020] S1. Establish energy coupling and energy transfer model

[0021] The present invention establishes a model for coupling and transferring energy information between multi-temperature heat sources during the fully coupled operation of the passenger compartment, power battery and motor electronic control thermal management requirements of pure electric vehicles. The model quantitatively analyzes the heat supply and demand laws under different working modes, providing a basis for subsequent system optimization. Specifically, the model takes into account the different heat supply and demand laws of batteries, motors and passenger compartments under charging conditions, driving conditions and environmental conditions. For example, during driving under high temperature conditions, there may be a dual demand for passenger compartment cooling and battery cooling, that is, a battery passenger compartment dual cooling mode. In addition, the working modes also include single battery cooling, single battery heating, passenger compartment single cooling, passenger compartment single heating, and more complex battery passenger compartment dual cooling, battery passenger compartment dual heating motor waste heat recovery, battery cooling passenger compartment heating motor waste heat recovery and other multi-temperature heat source heat transfer modes. These modes cover situations ranging from single cooling or heating requirements to complex multi-heat source collaborative work, ensuring that the thermal management system of pure electric vehicles can effectively manage and optimize energy flow under different environmental and operating conditions to meet the needs of passenger comfort, battery performance and motor efficiency.

[0022] S2. Abstract the components in the thermal management system into network nodes, and then connect the network nodes based on different actual situations to form different information transmission networks, construct an undirected information structure graph based on the information transmission network, and construct a directed information structure graph based on the information transmission network and the energy information coupling and transmission model established in step S1, and describe the information transmission relationship between the components based on the undirected information structure graph and the directed information structure graph.

[0023] The components in the thermal management system (such as heat exchangers, compressors, expansion valves, water pumps, etc.) are abstracted as network nodes, and then the network nodes are connected based on different actual situations to form different information transfer networks. Based on the information transfer network, the undirected information structure diagram of the system is directly constructed, and the directed information structure diagram is constructed based on the energy information coupling and energy transfer model and the information transfer network, thereby providing a tool for revealing the information transfer relationship between components and facilitating the evaluation and optimization of the complex architecture of the integrated thermal management system.

[0024] The abstracted network node step only requires the participation of key components, such as one-way valves and other components that do not participate in the network node construction. The directed information structure diagram abstracted in a working mode may include a refrigerant circuit and a water pump circuit.

[0025] S3. Use the structural entropy algorithm to evaluate the total system order of different information transmission networks to quantify the transmission efficiency and accuracy of each information transmission network; for different information transmission networks, according to the timeliness entropy and quality entropy of the two vertices in the node pair, obtain the efficiency and accuracy of different paths, use efficiency and accuracy as path importance evaluation indicators, and identify important paths and non-important paths that need to be simplified in any information transmission network.

[0026] Please refer to Figure 1 , the present invention adopts the structural entropy method. The structural entropy adopts the principle of information entropy to map the shortest path of node pairs in the network and the metric indicators of nodes into the efficiency and accuracy of information transmission respectively. Through this method, the response speed of the network structure and the accuracy of information can be evaluated, and these factors can be combined to determine the overall order of the system. From the perspective of information flow, structural entropy is introduced into the evaluation of thermal management systems, the entropy values ​​of different structures are calculated, and the order of the architecture is quantified. The structural entropy is elaborated in detail below.

[0027] First, the timeliness microstate and quality microstate of the system and their realization probability must be determined. The calculation formulas are as follows: In the formula, i and j represent any two network nodes, p1(i,j) is the probability of realizing the time-sensitive microstate; L ij is the time-sensitive microstate, which represents the shortest path from i to j between any two nodes in the system; p2(i) is the probability of achieving the quality microstate; D in (i) represents the in-degree of a single node, D out (i) represents the out-degree of a single node, and the sum of the two represents the quality micro-state of a single node.

[0028] The temporal entropy is calculated based on the realization probability of the temporal microstate, and the calculation formula is as follows: H1(i,j)=-p1(i,j)log a p1(i,j) Where H1(i,j) is the temporal entropy; a=2.

[0029] This indicator can characterize the uncertainty of information transmission between two points.

[0030] The mass entropy is calculated based on the realization probability of the mass microstate, and the calculation formula is as follows: H2(i)=-p2(i)log a p2(i) Where H2(i) is the mass entropy; a=2.

[0031] This indicator can characterize the uncertainty of the accuracy of information received by a single vertex.

[0032] The total timeliness of the system is obtained by adding up the timeliness entropy of all vertices, which represents the uncertainty of the information transmission efficiency of the entire system. The calculation formula is as follows:

[0033] The total mass entropy of the system is obtained by adding up the mass entropies of all vertices, which represents the uncertainty of the accuracy of information transmission of the entire system. The calculation formula is as follows:

[0034] The maximum timeliness that a computing system can theoretically achieve can be calculated based on the defined timeliness microstate, as follows:

[0035] The maximum mass entropy that can be achieved theoretically can be calculated based on the defined mass microstate, as follows:

[0036] Calculate the system timeliness and define its theoretical range as [0,1]. The calculation formula is as follows: Where R1 is the system timeliness.

[0037] Calculate the system quality and define its theoretical range as [0,1]. The calculation formula is as follows: Where R2 is the system mass.

[0038] The larger the R1 value, the higher the information transmission efficiency of the system, and the larger the R2 value, the higher the information transmission accuracy of the system.

[0039] The timeliness and accuracy measures are added in a linear combination to calculate the overall order of the system. This order index comprehensively reflects the performance of the system in terms of information transmission efficiency and accuracy. The calculation formula is as follows: R=αR1+βR2 In the formula, R is the total order of the system; α and β represent the weight coefficients of the influence of the two attributes of system timeliness and system quality on the system. In the present invention, these two weights are set to 0.5 equally to balance their contributions to the total order. In order to adapt to the emphasis of different systems on these two attributes, their respective weights can also be adjusted.

[0040] The total order R of the system is calculated based on the energy coupling and energy transfer model, and the transmission efficiency and accuracy of different information transmission networks are judged based on the results of R1 and R2. For different information transmission networks, the efficiency and accuracy of different paths are obtained based on the timeliness entropy and quality entropy of the two vertices in the node pair. The efficiency and accuracy are used as path importance evaluation indicators to identify important paths and non-important paths that need to be simplified in any information transmission network, and optimize the energy transmission efficiency of the system from the architecture layer.

[0041] S4. Use the centrality algorithm to calculate the comprehensive centrality of each component in each information transmission network, use the comprehensive centrality of each component as an evaluation indicator of the role played by each component, determine the role of each component in information transmission in different information transmission networks, and identify the components with high comprehensive centrality in any information transmission network and the components with low comprehensive centrality that need to be simplified.

[0042] Based on structural entropy analysis, the centrality algorithm is used to evaluate the importance of components, calculate the centrality index of each integrated thermal management system component, determine the role of each component of the integrated thermal management system in information transmission, and identify the components with important information transmission functions in the integrated thermal management system.

[0043] Please refer to Figure 2 , the centrality algorithm belongs to the knowledge field of graph theory. Degree centrality, closeness centrality and betweenness centrality in graph theory can be used to evaluate the importance of each component in the system. Degree centrality measures the extensiveness of a node's connections with other nodes in the entire network, reflecting the number of direct connections between the node and other nodes in the system. The more connections a node has, the higher its degree centrality value, indicating that its core role in the network is more significant. The calculation formula is as follows: In the formula, C D (i) is the degree centrality of the node; D in (i) is the in-degree of the node; D out (i) is the out-degree of the node; N is the total number of nodes in the system structure network diagram.

[0044] Calculate the closeness centrality, which measures the centrality of a node in the network. It is calculated based on the inverse of the sum of the distances from the node to all other nodes. This metric combines the proximity of inbound and outbound nodes to measure the connectivity and information dissemination efficiency of the node in the entire network. The calculation formula is as follows: In the formula, i, j, k are arbitrary nodes; B in (i) and B out (i) are the number of inbound and outbound reachable vertices, respectively; L ji For B in (i) The shortest distance from vertex j to vertex i; L ik For vertex i to B out (i) The shortest distance corresponding to vertex k.

[0045] Calculate the betweenness centrality, which measures the criticality of a node in the network. It is determined by calculating the frequency of the node's appearance in the shortest path of all vertex pairs, reflecting its control over the flow of information. The calculation formula is as follows: In the formula, σ st is the number of shortest paths from vertex s to t that does not include vertex i; σ st (i) is σ st The number of paths passing through vertex i in .

[0046] Degree centrality, closeness centrality, and betweenness centrality together depict the relative importance of components in the electric vehicle thermal management system, evaluating their centrality in the entire system from multiple dimensions.

[0047] Calculate the standard deviation of degree centrality, closeness centrality, and betweenness centrality respectively.

[0048] Calculate C D The average value μ of (i) D , calculate the standard deviation of degree centrality σ D :

[0049] Calculate C C The average value μ of (i) C , calculate the standard deviation of closeness centrality σ C :

[0050] Calculate C B The average value μ of (i) B , calculate the standard deviation of betweenness centrality σ B :

[0051] Then the three standard deviations are scaled to the interval [0,1] using the Min-Max Normalization method. In the formula, σ′ D represents the normalized standard deviation of degree centrality, σ c ′ represents the normalized standard deviation of closeness centrality, σ′ B represents the normalized betweenness centrality standard deviation, max(σ) and min(σ) represent the maximum and minimum of the three standard deviation values, respectively.

[0052] The weighted average of the normalized standard deviations is calculated to obtain a comprehensive normalized centrality index. The present invention defines it as the comprehensive centrality σ: σ=ω D ·σ′ D +ω C ·σ′ C+ω B ·σ′ B In the formula, ω D ,ω C ,ω B are the weight coefficients of normalized degree centrality, closeness centrality, and betweenness centrality, respectively, and ω is taken D =0.3,ω C =0.4,ω B =0.3.

[0053] According to the calculation results of comprehensive centrality, all components are ranked, and the components with the highest centrality index are identified as those with important information transmission functions.

[0054] S5. Combine the order calculated by the structural entropy algorithm with the comprehensive centrality index calculated by the centrality algorithm to measure the complexity of different information transmission networks and evaluate different information transmission networks.

[0055] Combining the order degree calculated by the structural entropy and the comprehensive centrality index calculated by the centrality algorithm, the system complexity SC is calculated: SC=γ(1-R)+ησ Where SC is the system complexity, R is the system order, σ is the system comprehensive centrality, γ and η represent the weight coefficients of order and comprehensive centrality respectively, both of which are taken as 0.5 here.

[0056] S6. Based on the judgment results of step S3 and step S4, simplify non-important paths and components with low comprehensive centrality, and strengthen the connection between components with high comprehensive centrality to optimize the architecture of each information transmission network and improve the efficiency and stability of information transmission.

[0057] The system architecture evaluation and optimization method of the present invention not only takes into account the thermodynamic performance, but also comprehensively evaluates the information transmission efficiency and complexity of the system architecture. The application of structural entropy enables this method to quantify the order of different information transmission networks, identify bottlenecks and redundancies in information transmission, and thus provide a theoretical basis for the simplification and optimization of different information transmission networks. The centrality algorithm can help designers identify key components, strengthen the connection between components with high comprehensive centrality, and improve the overall performance and stability of the system by evaluating the importance of components in the information transmission network. Finally, the advantages and disadvantages of the architectures are measured according to the complexity of the system, and the optimized architecture is evaluated.

[0058] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Any equivalent changes or modifications made according to the essence of the present invention should be included in the protection scope of the present invention.

Claims

1. A system architecture evaluation and optimization method based on structural entropy and centrality algorithm, used to evaluate and optimize the integrated thermal management system architecture of electric vehicles, characterized in that: The main steps are as follows: S1. Establish a model for the energy information coupling and transfer between multi-temperature heat sources during the fully coupled operation of the passenger compartment, power battery and motor electronic control thermal management requirements of pure electric vehicles; S2. Abstract the components in the thermal management system into network nodes, and then connect the network nodes based on different actual situations to form different information transmission networks, construct an undirected information structure graph based on the information transmission network, construct a directed information structure graph based on the information transmission network and the energy information coupling and transmission model established in step S1, and describe the information transmission relationship between the components based on the undirected information structure graph and the directed information structure graph; S3. Use the structural entropy algorithm to evaluate the total order of different information transmission networks to quantify the transmission efficiency and accuracy of each information transmission network; for different information transmission networks, according to the timeliness entropy and quality entropy of the two vertices in the node pair, obtain the efficiency and accuracy of different paths, use efficiency and accuracy as path importance evaluation indicators, and identify important paths and non-important paths that need to be simplified in any information transmission network; S4. Use the centrality algorithm to calculate the comprehensive centrality of each component in each information transmission network, use the comprehensive centrality of each component as an evaluation indicator of the role played by each component, determine the role of each component in information transmission in different information transmission networks, and identify the components with high comprehensive centrality and the components with low comprehensive centrality that need to be simplified in any information transmission network; S5. Combine the order calculated by the structural entropy algorithm and the comprehensive centrality index calculated by the centrality algorithm to measure the complexity of different information transmission networks, so as to evaluate different information transmission networks; S6. Based on the judgment results of step S3 and step S4, simplify non-important paths and components with low comprehensive centrality, and strengthen the connection between components with high comprehensive centrality to optimize the architecture of each information transmission network and improve the efficiency and stability of information transmission.

2. According to claim 1, a system architecture evaluation and optimization method based on structural entropy and centrality algorithm is characterized in that: The centrality algorithms include degree centrality, closeness centrality, and betweenness centrality.

3. A system architecture evaluation and optimization method based on structural entropy and centrality algorithm according to claim 2, characterized in that: Degree centrality, closeness centrality, and betweenness centrality are used to calculate the standard deviation of degree centrality, closeness centrality, and betweenness centrality respectively, and then the three standard deviation values ​​are normalized. After that, the normalized standard deviation of degree centrality, closeness centrality, and betweenness centrality can be used to calculate the comprehensive centrality σ. According to the calculation results of the comprehensive centrality, all components are ranked, and the components with the highest centrality index are identified as the key components.

4. The system architecture evaluation and optimization method based on structural entropy and centrality algorithm according to claim 3 is characterized in that: The calculation formula of comprehensive centrality σ is: s = ω D ·s′ D +oh C ·s′ C +oh B ·s′ B In the formula, σ′ D represents the normalized standard deviation of degree centrality, σ c ′ represents the normalized standard deviation of closeness centrality, σ′ B represents the normalized standard deviation of betweenness centrality, ω D ,ω C ,ω B are the weight coefficients of normalized degree centrality, closeness centrality, and betweenness centrality, respectively, and ω is taken D =0.3,ω C =0.4,ω B =0.

3.

5. The system architecture evaluation and optimization method based on structural entropy and centrality algorithm according to claim 1 is characterized in that: Combining the order calculated by structural entropy and the comprehensive centrality index calculated by the centrality algorithm, the complexity SC of the information transmission network is calculated: SC=γ(1-R)+ησ Where SC is the system complexity, R is the total order of the system, σ is the comprehensive centrality, γ and η represent the weight coefficients of the total order of the system and the comprehensive centrality, respectively, and both are taken as 0.5.

Images