High-precision and stable rolling bearing fault diagnosis method

Abstract

The invention relates to a rolling bearing fault diagnosis method, in particular to a high-precision and stable rolling bearing fault diagnosis method, which adopts a fault diagnosis model combining a chaotic mapping strategy, a Cauchy variation strategy and a Levy flight strategy, the model performs population initialization by introducing the chaotic mapping strategy, the population diversity is remarkably improved, and the reliability of the model is improved. Therefore, the algorithm is helped to find a high-quality solution more quickly. In the search process, disturbance of the search process is increased by using a Cauchy variation strategy, and the risk of falling into local optimum is reduced. In addition, when iteration is finished each time, a Levy flight strategy is introduced to enable the model to occasionally perform long-distance exploration on the basis of short-distance search, and the search range is further expanded. By integrating the improvements, the rolling bearing fault diagnosis model is effectively prevented from falling into a local optimal solution, the stability of the model is improved, the classification precision is remarkably improved, a new thought and method are provided for the development of an intelligent fault diagnosis technology, and the rolling bearing fault diagnosis method has a wide application prospect.

Description

Technical Field

[0001] The invention relates to a rolling bearing fault diagnosis method, in particular to a rolling bearing fault diagnosis method with high precision and stability. Background Art

[0002] The health status of rolling bearings directly affects the safety and stable operation of mechanical equipment. Therefore, timely detection of faults is crucial. However, many current fault diagnosis algorithms face the problem of falling into local optimal solutions in practical applications, which seriously affects the classification accuracy. These algorithms usually rely on iterative optimization processes. In complex fault signals, they are prone to stay in suboptimal solutions due to improper initial parameter selection or insufficient search mechanisms, thus failing to fully explore the key features in the signal. For example, traditional optimization algorithms such as particle swarm optimization and genetic algorithms often experience premature convergence when processing multi-peak functions, and are unable to effectively explore a wider solution space. In fault diagnosis with the same data and parameters, the classification accuracy fluctuates greatly.

[0003] This limitation affects the stability and accuracy of the classification results in rolling bearing fault diagnosis, and may even lead to missed diagnosis or misdiagnosis, thereby increasing the risk of equipment downtime and safety hazards. Therefore, it is necessary to develop new optimization strategies to enhance the global search capability of the algorithm and reduce the negative impact of local optimal solutions on classification accuracy.

[0004] Most traditional fault diagnosis algorithms face the problems of improper initial parameter selection or insufficient search mechanism. These problems cause the search range of population members to gradually narrow as the iteration proceeds, and the exploration ability to significantly decrease. It is very easy to fall into the local optimal solution, which ultimately leads to large fluctuations in classification accuracy. The stability and accuracy of the classification results are therefore seriously affected. Summary of the invention

[0005] The purpose of the present invention is to provide a rolling bearing fault diagnosis method with high accuracy and stability. The present invention adopts a fault diagnosis model combining chaos mapping strategy, Cauchy mutation strategy and Levy flight strategy, aiming to integrate multiple optimization strategies to more effectively solve the problem of fault diagnosis algorithm falling into local optimality. The algorithm significantly improves the classification accuracy of rolling bearing faults by optimizing population initialization and improving search mechanism, providing new ideas and methods for the development of intelligent fault diagnosis technology.

[0006] The objective of the present invention is achieved through the following technical solutions: A rolling bearing fault diagnosis method with high accuracy and stability, the diagnosis method comprising the following process: Firstly, the fault diagnosis model combining the chaotic mapping strategy, the Cauchy mutation strategy and the Levy flight strategy is optimized by optimizing the algorithm at different stages during the random initialization of the population, the search process and after each iteration to enhance the uniformity of the initial distribution of the population. The Logistic chaotic mapping strategy of the present invention initializes the population by generating pseudo-random numbers, such as ( Figure 2 ); In order to expand the search range and avoid the algorithm from falling into the local optimum, the Cauchy distribution is used as the step factor and multiplied with the original displacement parameter in the search process to increase the disturbance of the fault diagnosis algorithm and enhance the ability of the fault diagnosis algorithm to jump out of the local optimal solution; the Levy flight strategy is introduced to increase the diversity of the population. The Levy flight strategy generates a set of random vectors with the same dimension as the population, and multiplies the vector element by element with the optimal member to form a new candidate member; then, the new member is compared with the original optimal member, and the better individual is retained as the new optimal member; in this way, the search range is expanded and the global optimization ability and search efficiency of the algorithm are improved.

[0007] The advantages and effects of the present invention are: The present invention adopts a fault diagnosis model that combines a chaotic mapping strategy, a Cauchy mutation strategy and a Levy flight strategy to effectively solve the local optimal problem. In a randomly initialized population, the model initializes the population by introducing a chaotic mapping strategy, which significantly improves the diversity of the population, thereby helping the algorithm to find a high-quality solution faster. During the search process, the Cauchy mutation strategy is used to increase the disturbance of the search process and reduce the risk of falling into a local optimal solution. In addition, at the end of each iteration, the introduction of the Levy flight strategy allows the model to occasionally perform long-distance exploration on the basis of short-distance search, further expanding the search range. Combining these improvements, the rolling bearing fault diagnosis model effectively avoids falling into a local optimal solution, improves the stability of the model, and significantly improves the classification accuracy, providing new ideas and methods for the development of intelligent fault diagnosis technology, and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 Flowchart for realizing the model of the present invention; Figure 2 This is a population diversity comparison chart of different population initialization functions of the present invention. DETAILED DESCRIPTION

[0009] The present invention will be described in detail below with reference to the embodiments shown in the accompanying drawings.

[0010] The present invention adopts a fault diagnosis model that combines the chaotic mapping strategy, the Cauchy mutation strategy and the Levy flight strategy. By optimizing the algorithm at different stages during the random initialization of the population, the search process and the end of each iteration, the uniformity of the initial distribution of the population is enhanced, and the diversity of the search mechanism and the ability to jump out of the local optimum are improved. By improving the balance between global search and local development, the model effectively improves the global convergence ability and classification accuracy of the diagnosis model, and significantly enhances the stability and reliability of the diagnosis results, providing an efficient solution for fault diagnosis of complex systems.

[0011] In the random initialization population, the rand function is usually used to initialize the population. Although this method has high randomness, the distribution of individuals in the solution space is often uneven, which may lead to a slow search speed and insufficient population diversity. Although this method has strong randomness, the distribution of individuals in the solution space is often not uniform enough, which may affect the search efficiency and lead to insufficient population diversity. To solve this problem, the present invention introduces the Logistic chaotic mapping strategy to initialize the population by generating pseudo-random numbers. Figure 2 The influence of different initialization functions on population diversity is demonstrated. By comparing the initial population distribution diagrams generated by the rand function and the chaotic mapping strategy, it can be seen that although the population generated by the chaotic mapping strategy is slightly less random, it is more evenly distributed and there is no individual overlap, which indicates that the chaotic mapping strategy has advantages in diversity and quality.

[0012] The Cauchy distribution is a continuous probability distribution with a heavy-tail characteristic, which means that the probability density of its tail is higher, which can generate more extreme value disturbances, enhance the diversity of the search process, and help the algorithm jump out of the local optimal solution. In order to expand the search range and avoid the algorithm from falling into the local optimal solution, during the search process, the Cauchy distribution is used as the step size factor and multiplied with the original displacement parameter to increase the disturbance of the fault diagnosis algorithm and enhance the ability of the fault diagnosis algorithm to jump out of the local optimal solution.

[0013] The Levy flight strategy is a random search method that mainly performs short-distance search and occasionally performs long-distance search. It is widely used in the field of population optimization algorithms. In fault diagnosis algorithms, as iterations proceed, the search range of population members gradually narrows, causing the algorithm to easily fall into the local optimal solution, affecting the global optimization effect. To this end, the Levy flight strategy is introduced to increase population diversity and significantly improve this problem. At the end of each iteration, if the state of the optimal member has not changed after several iterations, the Levy flight strategy is applied to the optimal member. The Levy flight strategy generates a set of random vectors with the same dimension as the population, multiplies the vector element by element with the optimal member, and forms a new candidate member. Subsequently, the new member is compared with the original optimal member, and the better individual is retained as the new optimal member. In this way, the search range is expanded, effectively avoiding the problem of the population falling into the local optimal solution too early, thereby improving the global optimization ability and search efficiency of the algorithm.

Claims

1. A rolling bearing fault diagnosis method with high precision and stability, characterized in that: The diagnostic method The process includes: Firstly, the fault diagnosis model of the chaotic mapping strategy, the Cauchy mutation strategy and the Levy flight strategy is combined, and the algorithm is optimized at different stages during the random initialization of the population, the search process and after each iteration to enhance the uniformity of the initial distribution of the population. The Logistic chaotic mapping strategy of the present invention initializes the population by generating pseudo-random numbers, as shown in (Figure 2). In order to expand the search range and avoid the algorithm from falling into the local optimum, the Cauchy distribution is used as the step factor and multiplied with the original displacement parameter in the search process to increase the disturbance of the fault diagnosis algorithm and enhance the ability of the fault diagnosis algorithm to jump out of the local optimal solution. The Levy flight strategy is introduced to increase the diversity of the population. The Levy flight strategy generates a set of random vectors with the same dimension as the population, and multiplies the vector element by element with the optimal member to form a new candidate member. Subsequently, the new member is compared with the original optimal member, and the better individual is retained as the new optimal member. In this way, the search range is expanded, and the global optimization ability and search efficiency of the algorithm are improved.

Images