Hydraulic equipment anomaly detection method and system based on artificial intelligence learning

Abstract

The invention discloses a hydraulic equipment anomaly detection method and system based on artificial intelligence learning, and relates to the technical field of industrial automation and intelligent monitoring, and the detection method comprises the specific steps: data collection and preprocessing: collecting data in real time from a pressure sensor, a temperature sensor and a vibration sensor on hydraulic equipment, according to the method, the fuzzy logic and sensor data fusion technology is adopted, complementarity of pressure, temperature and vibration multi-dimensional sensor data in the hydraulic equipment is fully utilized, the fuzzy logic carries out fuzzification processing on the data, uncertainty and fuzziness in the sensor data can be effectively processed, and the accuracy of the sensor data is improved. And meanwhile, the comprehensive perception capability of the system on the equipment state is further enhanced through fusion of multi-sensor data, so that the anomaly detection is more robust, and stable operation can be realized under complex and changeable working conditions.

Description

Technical Field

[0001] The present invention relates to the field of industrial automation and intelligent monitoring technology, and in particular to a method and system for detecting anomalies of hydraulic equipment based on artificial intelligence learning. Background Art

[0002] With the continuous development of industrial technology, hydraulic equipment, as an important transmission and control device, plays an indispensable role in various fields. However, in the long-term operation of hydraulic equipment, due to the influence of multiple factors such as wear, aging, and changes in the external environment, various faults and abnormal conditions are prone to occur. If these abnormal conditions are not discovered and handled in time, they will seriously affect the operating efficiency and safety of the equipment. Therefore, the abnormal detection and early warning technology of hydraulic equipment has become a research hotspot, aiming to monitor the operating status of the equipment in real time, discover and warn potential faults in time, so as to ensure the normal operation of the equipment.

[0003] In the abnormality detection of hydraulic equipment, traditional detection methods mostly rely on data monitoring and threshold judgment of a single sensor. This method is often difficult to fully reflect the operating status of the equipment and is easily affected by noise and abnormal value interference factors, resulting in insufficient accuracy and reliability of the detection results. In addition, although fuzzy logic has unique advantages in dealing with uncertainty and ambiguity, the fusion method of fuzzy logic and sensor data in traditional technology is relatively simple and lacks in-depth fusion mechanism, resulting in the construction of fuzzy reasoning rules is not precise enough, and it is difficult to accurately identify the complex abnormal status of the equipment. At the same time, the traditional system lacks continuous optimization and learning mechanism for fuzzy reasoning rules, and cannot adapt to the dynamic changes of hydraulic equipment under complex working conditions.

[0004] Therefore, in view of the deficiencies and shortcomings of the prior art, the present invention proposes a hydraulic equipment anomaly detection method and system based on artificial intelligence learning, which fuzzifies the data through fuzzy logic, effectively solves the uncertainty and ambiguity problems in sensor data, improves the accuracy of anomaly detection, and constructs a fuzzy reasoning rule library based on expert knowledge and historical data analysis, realizes accurate mapping and comprehensive judgment of real-time sensor data, enables the system to quickly identify potential fault hazards and trigger early warning mechanisms. Summary of the invention

[0005] The purpose of the present invention is to make up for the shortcomings of the prior art and to provide a hydraulic equipment anomaly detection method and system based on artificial intelligence learning. It realizes the deep integration and efficient utilization of fuzzy logic and sensor data through data acquisition and preprocessing, fuzzy set definition, fuzzy reasoning rule construction, fuzzy reasoning and anomaly judgment, and anomaly judgment and alarm steps. In particular, by introducing machine learning algorithms to optimize the fuzzy reasoning rule base, the system can continuously learn and adapt to the dynamic changes of hydraulic equipment, thereby improving the accuracy and robustness of anomaly detection.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: On the one hand, a method for detecting abnormalities of hydraulic equipment based on artificial intelligence learning, the specific steps of the detection method are:

[0007] S100, data collection and preprocessing: collect data from pressure sensors, temperature sensors, and vibration sensors on hydraulic equipment in real time, and perform preprocessing operations on the data;

[0008] S200, fuzzy set definition: for each sensor data type, define the corresponding fuzzy set, each fuzzy set describes the degree to which the data belongs to the set through a membership function;

[0009] S300, fuzzy reasoning rule construction: Based on expert knowledge and historical data, a series of fuzzy reasoning rules are constructed to define the status and abnormality type of the equipment under different combinations of sensor data fuzzy sets;

[0010] S400, fuzzy reasoning and abnormality judgment: using the fuzzy reasoning engine, according to the membership degree of each sensor data and the fuzzy reasoning rules, a comprehensive judgment is made to obtain the current operating status and abnormality type of the equipment;

[0011] S500, abnormality judgment and alarm: based on the fuzzy reasoning result, judge whether the hydraulic equipment is in an abnormal state. When it is judged that the abnormality triggers the alarm mechanism, notify the operator to execute the corresponding emergency treatment measures.

[0012] Further, the S100 includes:

[0013] S101, collecting raw data of pressure, temperature and vibration of hydraulic equipment in real time through pressure sensors, temperature sensors and vibration sensors, and storing the collected data in a database;

[0014] S102, denoising the collected data, checking and processing outliers in the data, smoothing the denoised and outlier-processed data, and standardizing the processed data according to the operating characteristics of the hydraulic equipment and data processing requirements.

[0015] Furthermore, the S200 includes:

[0016] S201, analyzing the physical meaning of each sensor data and its relationship with the equipment operation status, and performing statistical analysis based on historical data to understand the distribution characteristics and change trends of the data;

[0017] S202, according to the data characteristics and monitoring requirements, determine the number of fuzzy sets that need to be defined, give each fuzzy set a clear and easy-to-understand name, and determine the approximate range of each fuzzy set based on the results of physical and statistical property analysis;

[0018] S203, according to the distribution characteristics of the data and actual needs, the membership function model is used to calculate the degree to which the data element belongs to a fuzzy set, and the parameters of the function are determined by using expert experience and historical data;

[0019] S204, using simulated data to verify the accuracy of the model, and fine-tuning the parameters of the model based on the verification results.

[0020] Furthermore, the S203 calculates the degree to which the membership function model describes the data element belongs to a fuzzy set, and the mathematical expression is: Among them, μ is the mean of the data, and the calculation formula is: k is a tuning parameter, x is the sensor reading, when x = μ, When x approaches positive infinity, tanh(k(x-μ)) approaches 1, and f(x) approaches 0. When x approaches negative infinity, tanh(k(x-μ)) approaches -1, and f(x) approaches 1. As x changes from negative infinity to positive infinity, the value of f(x) gradually decreases from 1 to Then to 0, presenting a smooth transition.

[0021] Furthermore, the S300 includes:

[0022] S301, collect the in-depth understanding and experience summary of the equipment operation status and abnormal conditions from experts in the field of hydraulic equipment;

[0023] S302, analyzing the distribution characteristics and change rules of the historical operation data of the hydraulic equipment, identifying abnormal key features and patterns, and calculating the potential relationships and patterns between the sensor data through the correlation function;

[0024] S303, forming a basic rule framework based on expert knowledge and historical data analysis;

[0025] S304, based on the relationship between the device status and the data, the input conditions and output results of each rule are refined, and the initially constructed rules are verified using historical data;

[0026] S305, optimizing the rules according to the verification results, repeating the verification and optimization process until the rules achieve satisfactory performance.

[0027] Furthermore, the S302 uses the potential relationship and pattern between the multiple sensor data through the correlation function, assuming that the sensor data has two data sets X = {x1, x2, ..., x n} and Y={y1,y2,…,y n}, the formula of correlation function R(X,Y) is: When x i and i When the changing trends of x are very similar, i Increase y i also increases, and x i Decrease y i Also decreases, the numerator 2|x i y i |Will be close The value of R(X,Y) will be close to 1, and there is a strong positive correlation between the data. i and i When the trend of x is completely opposite, i Increase y i Decrease, numerator 2|x i y i | will be close to 0, the value of R(X,Y) will be close to 0, there is a negative correlation and weak correlation between the data, x i and i When there is no obvious pattern between them, the value of R(X,Y) will fluctuate between 0 and 1.

[0028] Furthermore, the S400 includes:

[0029] S401, collecting data from each sensor in real time, mapping the collected sensor data to the corresponding fuzzy set, and calculating its membership degree;

[0030] S402, using the membership degree of the fused multi-sensor data fuzzy set as input, matching it with the rules in the rule base, and judging the possibility of abnormality of the current state of the device through the fuzzy rule algorithm for the data combination that meets the rule conditions;

[0031] S403, performing a comprehensive evaluation based on the states and membership degrees obtained by fuzzy reasoning.

[0032] Furthermore, the S402 determines the possibility of abnormality in the current state of the device through a fuzzy rule algorithm, and assumes that the fuzzy set membership vector of the fused multi-sensor data is Among them, m i Represents the membership of the i-th sensor data, and defines the function To indicate the possibility of abnormality in the current state of the device, its value range is [0, 1]. The closer it is to 1, the greater the possibility of an abnormality in the device. The closer it is to 0, the smaller the possibility of an abnormality in the device. Suppose there are k rules R1, R2, …, R k , each rule R j There is a corresponding weight w j ,and The weight indicates the importance of the rule in judging the possibility of device abnormality. For each rule R j , define a function When the input membership vector Satisfy rule R j If the condition is The probability function of abnormality in the current state of the device for: The obtained fuzzy set membership vector of the fused multi-sensor data Substitute into the function In the function The calculated value determines the possibility of abnormality in the current state of the device.

[0033] Furthermore, the S500 includes:

[0034] S501, receiving a comprehensive evaluation result on the hydraulic equipment state, setting clear thresholds for different abnormal states, and determining that the state is abnormal if the membership degree of the abnormal state exceeds the corresponding threshold;

[0035] S502, once it is determined that the device is in an abnormal state, the alarm mechanism will be triggered immediately, and corresponding alarm information will be generated according to the type and severity of the abnormal state and transmitted to the operator;

[0036] S503, after receiving the alarm information, the operator immediately checks and confirms the alarm content, and executes corresponding emergency treatment measures according to the treatment strategy and suggestions provided in the alarm information.

[0037] On the other hand, a hydraulic equipment anomaly detection system based on artificial intelligence learning, the specific steps of the system are:

[0038] Data acquisition and preprocessing module: real-time acquisition of pressure, temperature, flow, vibration and other sensor data of hydraulic equipment, and data cleaning and denoising preprocessing operations;

[0039] Fuzzy set definition and data fuzzification module: According to the operating characteristics of the hydraulic equipment, different fuzzy sets are set for each sensor, and the pre-processed sensor data is mapped to the corresponding fuzzy set to realize the fuzzification of data;

[0040] Fuzzy reasoning rule base construction module: Based on expert knowledge and historical data, it defines the logical relationship between different sensor data fuzzy set combinations and equipment operation status abnormalities, and builds a complete fuzzy reasoning rule base;

[0041] Artificial intelligence learning optimization module: introduces machine learning to the fuzzy reasoning rule library, automatically adjusts and optimizes the fuzzy reasoning rules by continuously learning and analyzing new equipment operation data;

[0042] Abnormal detection and early warning module: Based on the fuzzy reasoning rule base and the fuzzy set of current sensor data, real-time reasoning and judgment are performed. Once an abnormal state is identified, the early warning mechanism is immediately triggered to notify relevant personnel to take corresponding measures.

[0043] Compared with the prior art, the hydraulic equipment abnormality detection method and system based on artificial intelligence learning has the following beneficial effects:

[0044] 1. The present invention adopts fuzzy logic and sensor data fusion technology, making full use of the complementarity of multi-dimensional sensor data of pressure, temperature and vibration in hydraulic equipment. Fuzzy logic can effectively deal with the uncertainty and ambiguity in sensor data by fuzzifying the data, thereby improving the accuracy of anomaly detection. At the same time, the fusion of multi-sensor data further enhances the system's comprehensive perception of the equipment status, making anomaly detection more robust and able to operate stably under complex and changeable working conditions.

[0045] 2. The fuzzy reasoning rule base constructed by the present invention can accurately identify the abnormal state of hydraulic equipment based on expert knowledge and historical data analysis. By mapping real-time sensor data to corresponding fuzzy sets and using the fuzzy reasoning engine for comprehensive judgment, the system can quickly identify potential fault hazards and trigger the early warning mechanism in time. This early warning method based on fuzzy logic not only improves the timeliness of fault warning, but also improves the accuracy of early warning by refining abnormal types and setting clear thresholds, providing operators with more reliable decision support.

[0046] Other advantages, objectives and features of the present invention will be set forth in part in the following description and, in part, will be apparent to those skilled in the art based on an examination of the following or may be taught from the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0048] Figure 1 It is a flow chart of a method for detecting abnormality of hydraulic equipment based on artificial intelligence learning;

[0049] Figure 2 This is a flow chart of a hydraulic equipment anomaly detection system based on artificial intelligence learning. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. 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.

[0051] Embodiment 1:

[0052] A method for detecting anomalies of hydraulic equipment based on artificial intelligence learning comprises the following steps: data collection and preprocessing, fuzzy set definition, fuzzy reasoning rule construction, fuzzy reasoning and anomaly determination, anomaly determination and alarm.

[0053] First, enter the data acquisition and preprocessing stage (S100), install high-precision pressure sensors, temperature sensors and vibration sensors at key parts of the hydraulic equipment, and obtain the original data of pressure, temperature and vibration from the above sensors in real time through the data acquisition system. The collected original data is stored in a database for subsequent preprocessing and analysis. The original data stored in the database is denoised to remove noise signals to improve the signal-to-noise ratio of the data. After removing the noise signals, check whether there are abnormal values ​​in the data to avoid interference with subsequent analysis. According to the operating characteristics of the hydraulic equipment and data processing requirements, the smoothed data is standardized to eliminate the dimensional differences and numerical range differences between different sensor data, laying the foundation for subsequent data fusion and fuzzy processing.

[0054] Next, enter the fuzzy set definition stage (S200), conduct a detailed physical meaning analysis of the data collected by each sensor, clarify how these data reflect the operating status of the hydraulic equipment, and conduct statistical analysis based on the long-term operating history data of the equipment to understand the distribution characteristics and changing trends of the data, determine the number of fuzzy sets that need to be defined, and give each fuzzy set a name. Based on the analysis results of physical and statistical characteristics, further determine the specific scope of each fuzzy set. Based on the distribution characteristics and actual needs of the data, calculate the degree to which the membership describes the data element belongs to a fuzzy set through the membership function model, and use expert experience and historical data to determine the parameters of the model. In order to ensure the accuracy of the model, use simulation data to verify it, and fine-tune the parameters of the model based on the verification results until satisfactory performance is achieved.

[0055] Subsequently, the fuzzy reasoning rule construction stage (S300) is entered to collect experts' in-depth understanding and experience summary of the operating status and abnormal conditions of hydraulic equipment. The historical data accumulated from the long-term operation of hydraulic equipment is used to conduct in-depth statistical analysis to identify key features and patterns in the data that can indicate equipment abnormalities. The correlation function is introduced to explore the potential relationships and patterns between multiple sensor data, further revealing the interdependence and mutual influence between data. Based on the results of expert knowledge and historical data analysis, the basic framework of fuzzy reasoning rules is constructed. According to the relationship between equipment status and data, on the basis of the rule framework, the input conditions and output results of each rule are further refined. The historical data is used to verify the initially constructed rules. By simulating the equipment operating status and abnormal scenarios, the accuracy and reliability of the rules in identifying abnormal conditions are evaluated. According to the verification results, the rules are optimized and adjusted as necessary, and the verification and optimization process is repeated until the rule base achieves satisfactory performance.

[0056] Then, the fuzzy reasoning and abnormality judgment stage (S400) is entered. Real-time data is collected from the pressure sensors, temperature sensors, and vibration sensors on the hydraulic equipment. These collected sensor data are mapped to the previously defined fuzzy sets. During the mapping process, the membership of each data point to each fuzzy set is calculated. After the data mapping is completed, the fused fuzzy set membership of the multi-sensor data is used as input to match the rules in the constructed fuzzy reasoning rule library. For each rule, it is checked whether the current membership vector meets the conditions of the rule. If so, the possibility of abnormality in the current state of the equipment is calculated through the fuzzy rule algorithm, and a comprehensive evaluation is performed based on the membership of each state and abnormal state obtained by fuzzy reasoning.

[0057] Finally, the abnormality judgment and alarm stage (S500) is entered, and the comprehensive evaluation results of the hydraulic equipment are received. The thresholds of different abnormal states are comprehensively determined based on historical data, expert experience and the safe operation standards of the equipment. When the degree of membership of a certain abnormal state exceeds its corresponding threshold, the state is judged to be abnormal, and the alarm mechanism will be triggered immediately. According to the type and severity of the abnormal state, the corresponding alarm information is generated. The alarm information is transmitted to the operator through the preset communication channel. After receiving the alarm information, the operator immediately checks and confirms the authenticity and accuracy of the alarm content. Once the alarm content is confirmed to be correct, the operator should perform corresponding emergency treatment measures according to the processing strategies and suggestions provided in the alarm information.

[0058] Embodiment 2:

[0059] A method for anomaly detection of hydraulic equipment based on artificial intelligence learning places special emphasis on the application of fuzzy reasoning:

[0060] Analyze the data collected by each sensor to understand how these data reflect the operating status of the hydraulic equipment. Based on historical data, conduct statistical analysis to understand the distribution characteristics of the data. According to the data characteristics and monitoring needs, determine the number of fuzzy sets that need to be defined, and give each fuzzy set a clear and easy-to-understand name for subsequent operation and understanding. Based on the results of the physical and statistical characteristics analysis, determine the approximate range of each fuzzy set. These ranges should be able to cover all possible values ​​of the data while distinguishing different states and conditions. According to the distribution characteristics of the data and actual needs, the membership degree is used to define the fuzzy sets.

[0061] The function model calculates the degree of membership to describe the degree to which a data element belongs to a fuzzy set. The mathematical expression is: Among them, μ is the mean of the data, and the calculation formula is: k is a tuning parameter, x is the sensor reading, when x = μ, When x approaches positive infinity, tanh(k(x-μ)) approaches 1, and f(x) approaches 0. When x approaches negative infinity, tanh(k(x-μ)) approaches -1, and f(x) approaches 1. As x changes from negative infinity to positive infinity, the value of f(x) gradually decreases from 1 to Then to 0, presenting a smooth transition, expert experience and historical data are used to determine the parameters of the function, simulated data are used to calculate the membership to verify the accuracy of the model, and based on the verification results, the parameters of the model are fine-tuned, and the verification and fine-tuning process is repeated until the model reaches a satisfactory performance level.

[0062] Collect the in-depth understanding and experience summary of the equipment operation status and abnormal conditions of experts in the field of hydraulic equipment, extract key sensor data from the operation history database of hydraulic equipment, and use the potential relationship and pattern between multiple sensor data through the correlation function. Assume that the sensor data has two data sets X = {x1, x2, ..., x n} and Y={y1,y2,…,y n}, the formula of correlation function R(X,Y) is: When x i and i When the changing trends of x are very similar, i Increase y i also increases, and x i Decrease y i Also decreases, the numerator 2|x i y i |Will be close The value of R(X,Y) will be close to 1, and there is a strong positive correlation between the data. i and i When the trend of x is completely opposite, i Increase y i Decrease, numerator 2|x i y i | will be close to 0, the value of R(X,Y) will be close to 0, there is a negative correlation and weak correlation between the data, x i and i When there is no obvious pattern between them, the value of R(X,Y) will fluctuate between 0 and 1. Based on the results of expert knowledge and historical data analysis, a series of rule combinations are listed to construct the basic framework of fuzzy reasoning rules. According to the relationship between equipment status and data, on the basis of the rule framework, the input conditions and output results of each rule are further refined. The historical data are used to verify the preliminary constructed rules to check their accuracy and effectiveness in identifying abnormal equipment status. According to the verification results, the rules are adjusted and optimized as necessary. The verification and optimization process is repeated until the rule base achieves satisfactory performance and can accurately and timely identify abnormal conditions of hydraulic equipment.

[0063] Collect data from pressure sensors, temperature sensors, vibration sensors and other sensors in real time, map the collected sensor data to the previously defined fuzzy set, and calculate the membership of each data point to the corresponding fuzzy set. The membership describes the degree to which a data point belongs to a fuzzy set and is a basic concept in fuzzy set theory. Suppose the fuzzy set membership vector of the fused multi-sensor data is Among them, m iRepresents the membership of the i-th sensor data, matches the calculated membership vector with the rules in the fuzzy inference rule base, and defines the function To indicate the possibility of abnormality in the current state of the device, its value range is [0, 1]. The closer it is to 1, the greater the possibility of an abnormality in the device. The closer it is to 0, the smaller the possibility of an abnormality in the device. Suppose there are k rules R1, R2, …, R k , each rule R j There is a corresponding weight w j ,and The weight indicates the importance of the rule in judging the possibility of device abnormality. For each rule R j , define a function When the input membership vector Satisfy rule R j If the condition is The probability function of abnormality in the current state of the device for: The obtained fuzzy set membership vector of the fused multi-sensor data is Substitute into the function In the function The calculated value determines the possibility of abnormality in the current state of the equipment. A comprehensive evaluation is performed based on the membership of each state and abnormal state obtained by fuzzy reasoning. Based on the results of the comprehensive evaluation, the current operating state and abnormality type of the equipment are determined. If the equipment is in an abnormal state, the next step of abnormal judgment and alarm process is entered.

[0064] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A method for detecting abnormality of hydraulic equipment based on artificial intelligence learning, characterized in that: The specific steps of this detection method are: S100, data collection and preprocessing: collect data from pressure sensors, temperature sensors, and vibration sensors on hydraulic equipment in real time, and perform preprocessing operations on the data; S200, fuzzy set definition: for each sensor data type, define the corresponding fuzzy set, each fuzzy set describes the degree to which the data belongs to the set through a membership function; S300, fuzzy reasoning rule construction: Based on expert knowledge and historical data, a series of fuzzy reasoning rules are constructed to define the status and abnormality type of the equipment under different combinations of sensor data fuzzy sets; S400, fuzzy reasoning and abnormality judgment: using the fuzzy reasoning engine, according to the membership degree of each sensor data and the fuzzy reasoning rules, a comprehensive judgment is made to obtain the current operating status and abnormality type of the equipment; S500, abnormality judgment and alarm: based on the fuzzy reasoning result, judge whether the hydraulic equipment is in an abnormal state. When it is judged that the abnormality triggers the alarm mechanism, notify the operator to execute the corresponding emergency treatment measures.

2. According to claim 1, a method for detecting abnormalities in hydraulic equipment based on artificial intelligence learning is characterized in that: The S100 includes: S101, collecting raw data of pressure, temperature and vibration of hydraulic equipment in real time through pressure sensors, temperature sensors and vibration sensors, and storing the collected data in a database; S102, denoising the collected data, checking and processing outliers in the data, smoothing the denoised and outlier-processed data, and standardizing the processed data according to the operating characteristics of the hydraulic equipment and data processing requirements.

3. According to claim 1, a method for detecting abnormalities in hydraulic equipment based on artificial intelligence learning is characterized in that: The S200 includes: S201, analyzing the physical meaning of each sensor data and its relationship with the equipment operation status, and performing statistical analysis based on historical data to understand the distribution characteristics and change trends of the data; S202, according to the data characteristics and monitoring requirements, determine the number of fuzzy sets that need to be defined, give each fuzzy set a clear and easy-to-understand name, and determine the approximate range of each fuzzy set based on the results of physical and statistical property analysis; S203, according to the distribution characteristics of the data and actual needs, the membership function model is used to calculate the degree to which the data element belongs to a fuzzy set, and the parameters of the function are determined by using expert experience and historical data; S204, using simulated data to verify the accuracy of the model, and fine-tuning the parameters of the model based on the verification results.

4. According to claim 3, a method for detecting abnormalities in hydraulic equipment based on artificial intelligence learning is characterized in that: The S203 calculates the degree to which the membership function model describes the data element belongs to a fuzzy set, and the mathematical expression is: Among them, μ is the mean of the data, and the calculation formula is: k is a tuning parameter, x is the sensor reading, when x = μ, f When x approaches positive infinity, tanh(k(x-μ)) approaches 1, and f(x) approaches 0. When x approaches negative infinity, tanh(k(x-μ)) approaches -1, and f(x) approaches 1. As x changes from negative infinity to positive infinity, the value of f(x) gradually decreases from 1 to Then to 0, presenting a smooth transition.

5. The method for detecting abnormality of hydraulic equipment based on artificial intelligence learning according to claim 1 is characterized in that: The S300 includes: S301, collect the in-depth understanding and experience summary of the equipment operation status and abnormal conditions from experts in the field of hydraulic equipment; S302, analyzing the distribution characteristics and change rules of the historical operation data of the hydraulic equipment, identifying abnormal key features and patterns, and calculating the potential relationships and patterns between the sensor data through the correlation function; S303, forming a basic rule framework based on expert knowledge and historical data analysis; S304, based on the relationship between the device status and the data, the input conditions and output results of each rule are refined, and the initially constructed rules are verified using historical data; S305, optimizing the rules according to the verification results, repeating the verification and optimization process until the rules achieve satisfactory performance.

6. The method for detecting abnormality of hydraulic equipment based on artificial intelligence learning according to claim 5 is characterized in that: In step S302, the potential relationship and pattern between multiple sensor data are obtained by using the correlation function. Assume that the sensor data has two data sets X = {x1, x2, ..., x n } and Y={y1,y2,…,y n }, the formula of correlation function R(X,Y) is: When x i and i When the changing trends of x are very similar, i Increase y i also increases, and x i Decrease y i Also decreases, the numerator 2|x i y i |Will be close The value of R(X,Y) will be close to 1, and there is a strong positive correlation between the data. i and i When the trend of x is completely opposite, i Increase y i Decrease, numerator 2|x i y i | will be close to 0, the value of R(X,Y) will be close to 0, there is a negative correlation and weak correlation between the data, x i and i When there is no obvious pattern between them, the value of R(X,Y) will fluctuate between 0 and 1.

7. The method for detecting abnormality of hydraulic equipment based on artificial intelligence learning according to claim 1, characterized in that: The S400 includes: S401, collecting data from each sensor in real time, mapping the collected sensor data to the corresponding fuzzy set, and calculating its membership degree; S402, using the membership degree of the fused multi-sensor data fuzzy set as input, matching it with the rules in the rule base, and judging the possibility of abnormality of the current state of the device through the fuzzy rule algorithm for the data combination that meets the rule conditions; S403, performing a comprehensive evaluation based on the states and membership degrees obtained by fuzzy reasoning.

8. The method for detecting abnormality of hydraulic equipment based on artificial intelligence learning according to claim 7 is characterized in that: S402 determines the possibility of abnormality in the current state of the device through a fuzzy rule algorithm. Suppose the fuzzy set membership vector of the fused multi-sensor data is Among them, m i Represents the membership of the i-th sensor data, and defines the function To indicate the possibility of abnormality in the current state of the device, its value range is [0, 1]. The closer it is to 1, the greater the possibility of an abnormality in the device. The closer it is to 0, the smaller the possibility of an abnormality in the device. Suppose there are k rules R1, R2, …, R k , each rule R j There is a corresponding weight w j ,and The weight indicates the importance of the rule in judging the possibility of device abnormality. For each rule R j , define a function When the input membership vector Satisfy rule R j If the condition is The probability function of abnormality in the current state of the device for: The obtained fuzzy set membership vector of the fused multi-sensor data is Substitute into the function In the function The calculated value determines the possibility of abnormality in the current state of the device.

9. The method for detecting abnormality of hydraulic equipment based on artificial intelligence learning according to claim 1, characterized in that: The S500 includes: S501, receiving a comprehensive evaluation result on the hydraulic equipment state, setting clear thresholds for different abnormal states, and determining that the state is abnormal if the membership degree of the abnormal state exceeds the corresponding threshold; S502, once it is determined that the device is in an abnormal state, the alarm mechanism will be triggered immediately, and corresponding alarm information will be generated according to the type and severity of the abnormal state and transmitted to the operator; S503, after receiving the alarm information, the operator immediately checks and confirms the alarm content, and executes corresponding emergency treatment measures according to the treatment strategy and suggestions provided in the alarm information.

10. A hydraulic equipment anomaly detection system based on artificial intelligence learning, characterized in that: The specific steps of the system are: Data acquisition and preprocessing module: real-time acquisition of pressure, temperature, flow, vibration and other sensor data of hydraulic equipment, and data cleaning and denoising preprocessing operations; Fuzzy set definition and data fuzzification module: According to the operating characteristics of the hydraulic equipment, different fuzzy sets are set for each sensor, and the pre-processed sensor data is mapped to the corresponding fuzzy set to realize the fuzzification of data; Fuzzy reasoning rule base construction module: Based on expert knowledge and historical data, it defines the logical relationship between different sensor data fuzzy set combinations and equipment operation status abnormalities, and builds a complete fuzzy reasoning rule base; Artificial intelligence learning optimization module: introduces machine learning to the fuzzy reasoning rule library, automatically adjusts and optimizes the fuzzy reasoning rules by continuously learning and analyzing new equipment operation data; Abnormal detection and early warning module: Based on the fuzzy reasoning rule base and the fuzzy set of current sensor data, real-time reasoning and judgment are performed. Once an abnormal state is identified, the early warning mechanism is immediately triggered to notify relevant personnel to take corresponding measures.

Images