Automobile energy consumption data visualization method and system

Abstract

The invention relates to the technical field of data processing, in particular to an automobile energy consumption data visualization method and system. The method comprises the following steps: collecting all attribute values in a preset time period to form an attribute vector; obtaining an isolated tree of the attribute vector, and obtaining the importance degree of the attribute based on the correlation; obtaining the key degree of each attribute by combining the difference between the abnormal data of each attribute and all the attribute values of the attribute based on the number of times that the attributes are abnormal in the historical data; calculating the balance difference degree of the nodes of each isolated tree, and weighting the difference between the node balance difference degree and the same-layer node balance difference degree by taking the distance from each node to a root node as a weight to obtain the structural symmetry of the isolated tree; obtaining the weight of the isolated tree based on the key degree, the importance degree and the structural symmetry, obtaining the abnormal degree in combination with an attribute vector, and deleting the vector; and then constructing a knowledge graph to realize energy consumption data visualization. The reliability of the visualized automobile energy consumption knowledge graph is improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method and system for visualizing automobile energy consumption data. Background Art

[0002] In the context of big data on automotive energy consumption, the application of knowledge graphs has become crucial. By integrating and analyzing massive amounts of data from all aspects of the energy-consuming industry, including production, distribution, consumption, as well as policies, markets, and the environment, knowledge graphs provide decision makers with a powerful analytical and predictive platform. Leveraging natural language processing and machine learning technologies, knowledge graphs automatically identify and extract entities, relationships, and attributes from data, building a structured knowledge base that makes it easier to discover patterns, regularities, and trends within the data. Furthermore, the visualization capabilities of knowledge graphs present complex data analysis results in an intuitive and easy-to-understand manner, greatly enhancing transparency and engagement in decision-making. This not only supports automotive companies in making accurate decisions in complex market environments, but also dynamically updates and maintains the knowledge base to respond to changes in data sources, ensuring data timeliness and accuracy.

[0003] Before visualizing energy consumption data, data preprocessing is required, such as performing anomaly detection on the data. However, the existing data is being processed. When using isolation forests for data anomaly detection, the weights given to each generated tree are the same. However, in reality, the anomaly detection results of each tree generated are inconsistent. Therefore, when using the existing isolation forest algorithm for anomaly detection, the final detection results may be unsatisfactory. Therefore, it is necessary to analyze each randomly generated tree to avoid the above problems. Summary of the Invention

[0004] In order to solve the technical problem of poor anomaly detection effect, this application provides a method and system for visualizing automobile energy consumption data. The technical solutions adopted are as follows:

[0005] In a first aspect, the present application proposes a method for visualizing automobile energy consumption data, the method comprising the following steps:

[0006] Counting the attribute values ​​of different attributes within a preset time and forming an attribute vector from all attribute values ​​collected at each moment, wherein the attributes include the energy consumption of the vehicle;

[0007] All attribute vectors are processed through isolation forest to obtain several isolated trees; the importance of each attribute is obtained according to the correlation between different attributes in all attribute vectors;

[0008] Count the number of times each attribute in the historical data appears abnormal, and use the ratio of this number to the total number of collections for each attribute in the historical data as the probability of each attribute appearing abnormal. Obtain the degree of abnormality of each abnormal data in each attribute based on the difference between the attribute value of each abnormal data in each attribute in the historical data and the attribute value mean of each attribute in the historical data. Obtain the criticality of each attribute based on the degree of abnormality of all abnormal data in each attribute in the historical data and the probability of each attribute appearing abnormal in the historical data.

[0009] For each isolated tree, the difference between the left subtree depth and the right subtree depth of each node in the isolated tree is calculated to obtain the balanced difference of each node; using the distance of each node from the root node as the weight, the balanced difference of each node and the difference between the balanced difference of the node and the nodes on the same layer are weighted to obtain the structural symmetry of the isolated tree;

[0010] Calculate the attribute weight of each attribute based on its criticality and importance, and obtain the weight of the isolated tree based on the attribute weight and the structural symmetry of the isolated tree; use the weight of each isolated tree to weight the attribute value of each attribute vector in this isolated tree to obtain the abnormality of each attribute vector; prune vectors based on the abnormality degree;

[0011] The attribute vectors after excluding abnormal data are constructed into a knowledge graph, and the visualization of automobile energy consumption data is obtained based on the constructed knowledge graph.

[0012] In the above scheme, this application analyzes the obtained isolated trees, first divides the attributes according to the isolated trees, obtains the attribute weights of the data attributes corresponding to each isolated tree, and analyzes the structure of the isolated trees at the same time to obtain the final weights of the isolated trees. It greatly considers the different importance of different attributes to the data, and the problem of unreliable data detection results caused by the isolated tree structure, increases the reliability and accuracy of the isolated forest anomaly detection results, improves the quality of the energy consumption data after preprocessing, and improves the reliability of the knowledge graph of the energy consumption data after visualization.

[0013] In one embodiment, the attributes include the vehicle's driving distance, the vehicle's energy consumption, the vehicle's average speed, the average temperature of the external environment, the vehicle's altitude change value, and the external road congestion time; the knowledge graph of energy consumption data is updated once every preset time, and the attribute values ​​between the two knowledge graph updates are collected.

[0014] In one embodiment, the isolation tree contains only one attribute, and a leaf node of the isolation tree is an attribute value of the corresponding attribute in the attribute vector.

[0015] In one embodiment, the method for obtaining the importance of each attribute based on the correlation between different attributes in all attribute vectors is:

[0016] Sort the attribute values ​​of the same attribute in all attribute vectors in chronological order to obtain the attribute sequence of each attribute; calculate the correlation between each attribute sequence and the remaining attribute sequences, calculate the mean of all correlations corresponding to each attribute sequence, and use the mean as the importance of the attribute.

[0017] In one embodiment, the method for obtaining the abnormality degree of each abnormal data in each attribute based on the difference between the attribute value of each abnormal data in each attribute in the historical data and the attribute value mean of each attribute in the historical data is:

[0018] FC (j,r) Indicates the attribute value of the jth abnormal data corresponding to attribute r in the historical data, Bc r represents the mean of all attribute values ​​within attribute r in historical data, G (j,r) Indicates the abnormality degree of the jth abnormal data corresponding to attribute r in the historical data.

[0019] In one embodiment, the method for obtaining the criticality of each attribute based on the abnormality degree of all abnormal data in each attribute in the historical data and the probability of abnormality of each attribute in the historical data is:

[0020] The criticality of each attribute is negatively correlated with the abnormality of all abnormal data in each attribute, and positively correlated with the probability of abnormality in each attribute.

[0021] In one embodiment, the method for obtaining the structural symmetry of the isolated tree by weighting the balanced difference of each node and the difference between the balanced difference of the node and the nodes at the same layer using the distance of each node from the root node as a weight is:

[0022] H i represents the equilibrium difference of the i-th node, H (i,v) represents the difference in the degree of equilibrium difference between the i-th node and the v-th node on the same layer of the i-th node, L represents the depth of the isolated tree, S i represents the difference between the number of layers of the i-th node and the depth of the isolated tree, m represents the number of nodes in the isolated tree, exp() represents the exponential function with a natural constant as the base, and Hc represents the structural symmetry of the isolated tree.

[0023] In one embodiment, the method of calculating the attribute weight of each attribute according to the criticality and importance of each attribute, and obtaining the weight of the isolation tree based on the attribute weight and the structural symmetry of the isolation tree is:

[0024] The attribute weight is positively correlated with the criticality and importance of each attribute;

[0025] The weight of an isolated tree is positively correlated with the structural symmetry of the isolated tree and the attribute weight of the corresponding attribute of the isolated tree.

[0026] In one embodiment, the method of obtaining the abnormality degree of each attribute vector by weighting the attribute value of each isolated tree again by the weight of each isolated tree is:

[0027] Rd c represents the weight of the cth isolated tree, p (c,x) represents the attribute value of the xth attribute vector in the cth isolated tree, B represents the number of isolated trees, exp() represents the exponential function with a natural constant as the base, E x Indicates the abnormality of the x-th attribute vector.

[0028] In a second aspect, an embodiment of the present application also provides a vehicle energy consumption data visualization system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the system implements the steps of any one of the above-mentioned vehicle energy consumption data visualization methods.

[0029] The beneficial effects of this application are:

[0030] This application analyzes the obtained isolated trees, first divides the attributes according to the isolated trees, obtains the attribute weights of the data attributes corresponding to each isolated tree, and then analyzes according to the structure of the isolated trees to obtain the final weights of the isolated trees. It greatly considers the different importance of different attributes to the data and the problem of unreliable data detection results caused by the isolated tree structure, increases the reliability and accuracy of the isolation forest anomaly detection results, improves the quality of the energy consumption data after preprocessing, and improves the reliability of the knowledge graph of the visualized energy consumption data. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0032] Figure 1 A flow chart of a method for visualizing automobile energy consumption data provided in one embodiment of the present application. DETAILED DESCRIPTION

[0033] To further illustrate the technical means and effectiveness of this application's implementation of the intended invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effectiveness of a method and system for visualizing vehicle energy consumption data proposed in this application. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0034] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0035] An embodiment of a method and system for visualizing automobile energy consumption data:

[0036] The following describes in detail a specific method for visualizing automobile energy consumption data provided by this application with reference to the accompanying drawings.

[0037] See also Figure 1 , which shows a flow chart of a method for visualizing automobile energy consumption data provided by one embodiment of the present application, the method comprising the following steps:

[0038] Step S001: Count the attribute values ​​of different attributes within a preset time, and form an attribute vector from all attribute values ​​collected at each moment.

[0039] During the vehicle's operation phase, data is collected every half hour through onboard sensors and the map API. In this embodiment, the collected data includes the vehicle's distance traveled, energy consumption, average speed, average ambient temperature, altitude change, and duration of traffic congestion. Each attribute is considered an attribute, with data collected every half hour. The knowledge graph composed of historical data is updated every preset period.

[0040] In this embodiment, the knowledge graph is updated every 24 hours of operation of the car. The data collected every 24 hours of operation is pre-processed and added to the new knowledge graph to construct an updated knowledge graph to ensure the real-time visualization of the car's energy consumption.

[0041] The data of all types of attributes collected each time are combined into an attribute vector, and 48 sets of attribute vectors are obtained when the knowledge graph is updated twice.

[0042] At this point, all attribute vectors between knowledge graph updates are obtained.

[0043] Step S002: Process all attribute vectors through the isolation forest to obtain a number of isolated trees; obtain the importance of each attribute according to the correlation between different attributes in all attribute vectors.

[0044] All the attribute vectors obtained are processed using the isolation forest algorithm, and multiple isolated trees are obtained through the isolation forest algorithm, where each isolated tree corresponds to only one attribute, and the leaf node of the isolated tree is an attribute value of the corresponding attribute in the attribute vector.

[0045] For example, after all attribute vectors are processed through the isolation forest, the attribute of one of the isolated trees is the energy consumption of the vehicle. Then the value of each node in the isolated tree is the energy consumption value of the vehicle in each attribute vector.

[0046] The purpose of using the isolation forest algorithm in this application is to detect anomalies in the collected attribute vectors and eliminate abnormal data. Since the isolation forest randomly selects and constructs isolated trees based on different attributes, and there are differences between different attribute features, attribute weights are constructed for different isolated trees accordingly.

[0047] The attribute weight of an isolated tree is positively correlated with the importance of the attribute. That is, the greater the importance of the attribute, the more important the analysis result of the corresponding tree is, which means that the attribute weight of the tree is greater.

[0048] All attribute values ​​of each attribute in all attribute vectors constitute an attribute sequence. For example, all attribute values ​​corresponding to the energy consumption of the vehicle in all attribute vectors constitute its corresponding attribute sequence. The order of the sequence is time order.

[0049] Calculate the correlation between each attribute sequence and each of the other attribute sequences, average all the correlations corresponding to each attribute sequence, and use the average as the importance of the attribute. If the correlation between an attribute sequence and each of the other attribute sequences is poor, then the attribute's ability to characterize data features is stronger, and the corresponding attribute's importance is greater. In this embodiment, the correlation between attribute sequences is calculated using the Pearson correlation coefficient, which is a well-known technique and will not be described in detail in this embodiment. The implementer may use other correlation calculation methods.

[0050] At this point, the importance of each attribute is obtained.

[0051] Step S003: obtaining the probability of anomaly occurrence based on the number of anomalies of each attribute in the historical data, and obtaining the criticality of each attribute by combining the difference between the anomaly data of each attribute and all attribute values ​​of the attribute.

[0052] Collect all historical data of the vehicle operation stage and analyze the obtained historical data. Since the obtained data is acquired in real time, the update of the knowledge graph is also fixed. Due to changes in the external environment or aging of the instrument, there is a high possibility that some data will have abnormal test results. Therefore, the abnormal situation of the historical data can be analyzed based on this. The probability of abnormality of each attribute in the historical data is used as a reference, and then the importance of the obtained attributes is analyzed to obtain the criticality of each attribute.

[0053] In all historical data, the attribute data collected in all previous times are analyzed to obtain the number of times each attribute appears abnormal. The corresponding ratio of the number of times the attribute appears abnormal to the number of attribute data collected in the historical data is used as the probability of the attribute appearing abnormal.

[0054] And analyze the abnormal data obtained, analyze the abnormal degree of the abnormal data. If the data obtained may have a large number of abnormal times, but its abnormal degree is small, that is, the degree of deviation of the abnormal data from the normal value is small, and some data may deviate greatly from the normal data when they are abnormal, it is necessary to analyze the abnormal degree of the abnormal data. Based on the difference between the attribute value of each attribute with abnormality and the attribute value mean of the entire attribute in the historical data, calculate the abnormal degree of each abnormal data of each attribute. The expression is:

[0055] FC (j,r) Indicates the attribute value of the jth abnormal data corresponding to attribute r in the historical data, Bc r represents the mean of all attribute values ​​within attribute r in historical data, G (j,r) Indicates the abnormality degree of the jth abnormal data corresponding to attribute r in the historical data.

[0056] The greater the difference between the outlier and the mean, the greater the gap between the outlier and the normal value, that is, the greater the deviation between the outlier and the normal value, which means that the degree of abnormality of the abnormal data corresponding to the outlier is greater.

[0057] The criticality of each attribute is obtained based on the abnormality degree of all abnormal data in each attribute in the historical data and the probability of abnormality of each attribute in the historical data.

[0058] The criticality of each attribute is negatively correlated with the abnormality of all abnormal data in each attribute, and positively correlated with the probability of abnormality in each attribute.

[0059] It should be noted that positive correlation means that when one variable increases, the other variable also increases, and the two variables change in the same direction. When one variable changes from large to small or from small to large, the other variable also changes from large to small or from small to large; the specific relationship is determined by actual application and this application does not impose any special restrictions.

[0060] It should be noted that negative correlation means that when one variable increases, the other variable decreases accordingly, and the two variables change in opposite directions. When one variable changes from large to small or from small to large, the other variable also changes from small to large or from large to small. The specific relationship is determined by actual application and this application does not impose any special restrictions.

[0061] Preferably, in this embodiment, the expression for the criticality of each attribute is:

[0062] G (j,r) Indicates the abnormal degree of the jth abnormal data corresponding to attribute r in the historical data, P r represents the number of abnormal data in attribute r, F r Indicates the probability of anomaly of attribute r, exp() represents the exponential function with natural constant as the base, Tx r Indicates the criticality of each attribute.

[0063] The greater the probability that the attribute r to be sought is abnormal in historical data, and the greater the deviation between the abnormal value and the normal value, the greater the possibility that the attribute r is abnormal, and the more critical the attribute is.

[0064] At this point, the criticality of each attribute is obtained.

[0065] Step S004: Calculate the balanced difference of the nodes of each isolated tree, use the distance of each node from the root node as the weight, and weight the difference between the balanced difference of the nodes and the balanced difference of the nodes on the same layer to obtain the structural symmetry of the isolated tree.

[0066] For an isolated tree used for anomaly detection, the more asymmetric the tree structure is, the more it can reflect the characteristics of anomaly detection. That is, if the number of both sides of the entire tree is similar, then a relatively reliable anomaly detection result cannot be obtained based on the tree. Therefore, analysis can be performed based on the tree structure.

[0067] For each isolated tree, the depth of the left subtree and the depth of the right subtree of each node in the isolated tree are counted, and the balanced difference degree of each node is calculated based on the difference in the depth of the left and right subtrees. The specific expression is:

[0068] H i =|C (i,1) -C (i,2) |, C (i,1)Indicates the depth of the left subtree of the i-th node, C (i,2) Indicates the depth of the right subtree of the i-th node, H i Indicates the balance difference of the i-th node.

[0069] In an isolated tree, the greater the balance difference of each node, the more asymmetric the structure of each isolated tree. In addition, the more different the balance difference between two nodes on the same layer, the more different the two nodes are, and the more asymmetric the structure of the isolated tree is. That is, the difference between the balance difference of each node and any node on the same layer is calculated. The greater the difference, the more asymmetric the isolated tree structure is. The closer the distance to the root node of the isolated tree, the stronger the influence of each node. That is, the difference between the number of layers of each node and the depth of the isolated tree is calculated. The greater the difference, the farther away from the root node of the isolated tree. Therefore, based on the above content, the structural symmetry of each isolated tree is calculated as follows:

[0070] H i represents the equilibrium difference of the i-th node, H (i,v) represents the difference in the degree of equilibrium difference between the i-th node and the v-th node on the same layer of the i-th node, L represents the depth of the isolated tree, S i represents the difference between the number of layers of the i-th node and the depth of the isolated tree, m represents the number of nodes in the isolated tree, exp() represents the exponential function with a natural constant as the base, and Hc represents the structural symmetry of the isolated tree.

[0071] The farther each node is from the root node, the smaller its influence on the symmetry of the isolated tree. Therefore, the balance difference of each node is analyzed with its distance from the root node as the weight; the greater the balance difference of each node, the more asymmetric its isolated tree is. The greater the difference between the balance difference of each node and the nodes on the same layer, the greater the difference between the corresponding subtrees under the node, and the more asymmetric the isolated tree is.

[0072] At this point, the structural symmetry of each isolated tree is obtained.

[0073] Step S005: Obtain the weight of the isolated tree based on the criticality, importance and structural symmetry, obtain the abnormality degree in combination with the attribute vector, and delete the data.

[0074] Since each isolated tree corresponds to an attribute, the attribute weight of each attribute is calculated based on the criticality and importance of the attribute. The weight of each isolated tree is evaluated as a whole based on the attribute weight and the structural symmetry of the isolated tree. Anomaly detection is performed on each attribute vector based on the weight of the isolated tree.

[0075] Specifically, the attribute weight is positively correlated with the criticality and importance of each attribute.

[0076] Preferably, in this embodiment, the product of the criticality and importance of each attribute is used as the attribute weight of each attribute.

[0077] The weight of each isolated tree is obtained according to the structural symmetry of each isolated tree and the attribute weight of the attribute corresponding to the isolated tree.

[0078] The weight of an isolated tree is positively correlated with the structural symmetry of the isolated tree and the attribute weight of the corresponding attribute of the isolated tree.

[0079] Preferably, in this embodiment, the product of the structural symmetry of each isolated tree and the attribute weight of the attribute corresponding to the isolated tree is used as the weight of the isolated tree.

[0080] This gives us the weight of each isolated tree. When using the Isolation Forest algorithm for anomaly detection, the closer a leaf node is to the root of the isolated tree, the more abnormal it is. Each leaf node corresponds to an attribute vector. The degree of abnormality for each attribute vector is calculated by combining the abnormalities corresponding to that attribute vector in all isolated trees and the weight of the isolated tree. The expression is:

[0081] Rd c represents the weight of the cth isolated tree, p (c,x) represents the attribute value of the xth attribute vector in the cth isolated tree, B represents the number of isolated trees, E x Indicates the abnormality of the x-th attribute vector.

[0082] The above method is used to obtain the abnormality degree of each attribute vector between each two knowledge graph updates. If the abnormality degree is greater than the preset abnormality threshold, it means that the attribute vector is an abnormal attribute vector, and the attribute vector is excluded and does not participate in the subsequent knowledge graph construction.

[0083] Step S006: After deleting the data, a knowledge graph is constructed to visualize the energy consumption data.

[0084] Through the above steps, the anomaly detection of the collected data is completed, and the abnormal data is cleaned up. The collected attribute data is processed by the named entity recognition technology to obtain entities, and then the relationships are extracted by the relationship extraction algorithm. The entities, attributes, and relationships are formed into triples, and the knowledge graph is constructed by the triples. The method of constructing the knowledge graph is a well-known technology and will not be elaborated here.

[0085] Based on the constructed knowledge graph, automobile energy consumption data can be visualized.

[0086] Based on the same inventive concept as the above method, an embodiment of the present invention also provides an automobile energy consumption data visualization system, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, the steps of any one of the above-mentioned automobile energy consumption data visualization methods are implemented.

[0087] It should be noted that the above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the scope of protection of the present application.

[0088] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A method for visualizing automobile energy consumption data, characterized in that: The method comprises the following steps: Counting the attribute values ​​of different attributes within a preset time and forming an attribute vector from all attribute values ​​collected at each moment, wherein the attributes include the energy consumption of the vehicle; All attribute vectors are processed through isolation forest to obtain several isolated trees; the importance of each attribute is obtained according to the correlation between different attributes in all attribute vectors; Count the number of times each attribute in the historical data appears abnormal, and use the ratio of this number to the total number of collections for each attribute in the historical data as the probability of each attribute appearing abnormal. Obtain the degree of abnormality of each abnormal data in each attribute based on the difference between the attribute value of each abnormal data in each attribute in the historical data and the attribute value mean of each attribute in the historical data. Obtain the criticality of each attribute based on the degree of abnormality of all abnormal data in each attribute in the historical data and the probability of each attribute appearing abnormal in the historical data. For each isolated tree, the difference between the left subtree depth and the right subtree depth of each node in the isolated tree is calculated to obtain the balanced difference of each node; using the distance of each node from the root node as the weight, the balanced difference of each node and the difference between the balanced difference of the node and the nodes on the same layer are weighted to obtain the structural symmetry of the isolated tree; Calculate the attribute weight of each attribute based on its criticality and importance, and obtain the weight of the isolated tree based on the attribute weight and the structural symmetry of the isolated tree; use the weight of each isolated tree to weight the attribute value of each attribute vector in this isolated tree to obtain the abnormality of each attribute vector; prune vectors based on the abnormality degree; The attribute vectors after excluding abnormal data are constructed into a knowledge graph, and the visualization of automobile energy consumption data is obtained based on the constructed knowledge graph.

2. The method for visualizing automobile energy consumption data according to claim 1, characterized in that: Attributes include the vehicle's distance traveled, the vehicle's energy consumption, the vehicle's average speed, the average temperature of the external environment, the vehicle's altitude change, and the duration of external road congestion; the knowledge graph of energy consumption data is updated once every preset time, and the attribute values ​​between the two knowledge graph updates are collected.

3. The method for visualizing automobile energy consumption data according to claim 1, wherein: The isolated tree contains only one attribute, and the leaf node of the isolated tree is an attribute value of the corresponding attribute in the attribute vector.

4. The method for visualizing automobile energy consumption data according to claim 1, wherein: The method for obtaining the importance of each attribute based on the correlation between different attributes in all attribute vectors is: Sort the attribute values ​​of the same attribute in all attribute vectors in chronological order to obtain the attribute sequence of each attribute; calculate the correlation between each attribute sequence and the remaining attribute sequences, calculate the mean of all correlations corresponding to each attribute sequence, and use the mean as the importance of the attribute.

5. The method for visualizing automobile energy consumption data according to claim 1, wherein: The method for obtaining the abnormality degree of each abnormal data in each attribute according to the difference between the attribute value of each abnormal data in each attribute in the historical data and the attribute value mean of each attribute in the historical data is: FC (j,r) Indicates the attribute value of the jth abnormal data corresponding to attribute r in the historical data, Bc r represents the mean of all attribute values ​​within attribute r in historical data, G (j,r) Indicates the abnormality degree of the jth abnormal data corresponding to attribute r in the historical data.

6. The method for visualizing automobile energy consumption data according to claim 1, wherein: The method for obtaining the criticality of each attribute based on the abnormality degree of all abnormal data in each attribute in the historical data and the probability of abnormality of each attribute in the historical data is as follows: The criticality of each attribute is negatively correlated with the abnormality of all abnormal data in each attribute, and positively correlated with the probability of abnormality in each attribute.

7. The method for visualizing automobile energy consumption data according to claim 1, characterized in that: The method for obtaining the structural symmetry of the isolated tree by weighting the distance of each node from the root node and the difference between the balanced difference of each node and the balanced difference between the node and the nodes at the same level is as follows: H i represents the equilibrium difference of the i-th node, H (i,v) represents the difference in the degree of equilibrium difference between the i-th node and the v-th node on the same layer of the i-th node, L represents the depth of the isolated tree, S i represents the difference between the number of layers of the i-th node and the depth of the isolated tree, m represents the number of nodes in the isolated tree, exp() represents the exponential function with a natural constant as the base, and Hc represents the structural symmetry of the isolated tree.

8. The method for visualizing automobile energy consumption data according to claim 1, wherein: The method of calculating the attribute weight of each attribute according to the criticality and importance of each attribute, and obtaining the weight of the isolation tree based on the attribute weight and the structural symmetry of the isolation tree is: The attribute weight is positively correlated with the criticality and importance of each attribute; The weight of an isolated tree is positively correlated with the structural symmetry of the isolated tree and the attribute weight of the corresponding attribute of the isolated tree.

9. The method for visualizing automobile energy consumption data according to claim 1, wherein: The method for obtaining the abnormality degree of each attribute vector by weighting the attribute value of each isolated tree by the weight of each isolated tree is as follows: Rd c represents the weight of the cth isolated tree, p (c,x) represents the attribute value of the xth attribute vector in the cth isolated tree, B represents the number of isolated trees, exp() represents the exponential function with a natural constant as the base, E x Indicates the abnormality of the x-th attribute vector.

10. A vehicle energy consumption data visualization system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the automobile energy consumption data visualization method as described in any one of claims 1 to 9 are implemented.

Images