Intelligent operation and maintenance scheduling method for wind power plant based on equipment residual life prediction

Abstract

The invention relates to the field of intelligent operation and maintenance scheduling, in particular to a wind power plant intelligent operation and maintenance scheduling method based on equipment residual life prediction. The method comprises the following steps: collecting and preprocessing multi-dimensional operation data, and predicting the residual life of a fan based on the preprocessed multi-dimensional operation data; based on the residual life of the fan, generating a fault risk coefficient of the fan through a fan fault risk assessment algorithm; and based on the fault risk coefficient of the fan, generating an optimal operation and maintenance scheduling scheme through an intelligent operation and maintenance scheduling optimization algorithm of the wind power plant. The problems that maintenance priority distribution is unreasonable, cascading failure or resource waste is possibly caused by neglecting operation relevance between fans, and an optimal balance point is difficult to find between maintenance cost and shutdown loss are solved.

Description

Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance scheduling, and in particular to an intelligent operation and maintenance scheduling method for wind farms based on equipment remaining life prediction. Background Technology

[0002] As the scale of wind farm construction continues to expand and installed capacity continues to grow, the complexity and cost of wind power equipment operation and maintenance have also increased significantly. Therefore, how to scientifically predict the remaining useful life (RUL) of equipment and conduct intelligent operation and maintenance scheduling based on this has become a key issue for efficient wind farm operation and cost optimization. Wind farm operation and maintenance mainly includes three modes: preventative maintenance, reactive maintenance, and predictive maintenance. Traditional preventative maintenance relies on fixed-cycle inspections and replacements, which can reduce the probability of sudden failures but often leads to over- or under-maintenance. Reactive maintenance involves repairing equipment after a failure occurs; although lower in cost, it may cause prolonged downtime and affect power generation efficiency. In contrast, predictive maintenance, through real-time monitoring of equipment status combined with data analysis and fault prediction technology, can take intervention measures before failures occur, thereby optimizing maintenance costs and downtime.

[0003] Intelligent operation and maintenance scheduling methods for wind farms based on equipment remaining lifetime prediction are an important direction for the wind power industry to move towards intelligent and digital transformation. By combining advanced RUL prediction technology and optimized scheduling algorithms, it can play an important role in reducing maintenance costs, reducing downtime, and improving power generation efficiency. With the continuous advancement of artificial intelligence and Internet of Things technologies, intelligent operation and maintenance scheduling based on equipment remaining lifetime prediction will be more widely used in the future, providing strong technical support for the sustainable development of wind farms. Summary of the Invention

[0004] This invention provides a wind farm intelligent operation and maintenance scheduling method based on equipment remaining life prediction, in order to solve the problems of unreasonable maintenance priority allocation due to the failure to consider the differences in the contribution of wind turbines to the overall power generation benefits of the wind farm; the problem that traditional operation and maintenance scheduling ignores the operational correlation between wind turbines, which may lead to cascading failures or resource waste; and the problem that it is difficult to find the optimal balance between maintenance costs and downtime losses.

[0005] The present invention provides a method for intelligent operation and maintenance scheduling of wind farms based on equipment remaining life prediction, comprising the following steps: S1. Collect and preprocess multidimensional operating data, and predict the remaining life of the wind turbine based on the preprocessed multidimensional operating data; based on the remaining life of the wind turbine, generate the failure risk coefficient of the wind turbine through the wind turbine failure risk assessment algorithm; S2. Based on the failure risk coefficient of the wind turbine, the optimal operation and maintenance scheduling scheme is generated through the wind farm intelligent operation and maintenance scheduling optimization algorithm.

[0006] Preferably, S1 specifically includes: Long Short-Term Memory (LSTM) networks are used to predict the operating time of wind turbines before they fail in the future, and the remaining lifespan of each wind turbine is output.

[0007] Preferably, S1 specifically includes: In the implementation of the wind turbine failure risk assessment algorithm, the importance weight of the wind turbine is calculated based on its historical power generation, and the failure risk coefficient of each wind turbine is generated by combining the remaining life of each wind turbine.

[0008] Preferably, S2 specifically includes: The intelligent operation and maintenance scheduling optimization algorithm for wind farms aims to minimize total cost while satisfying operation and maintenance resource constraints, and generates the optimal operation and maintenance scheduling scheme.

[0009] Preferably, S2 specifically includes: The total cost is obtained by summing the maintenance costs of all wind turbines and the adjusted downtime losses.

[0010] Preferably, S2 specifically includes: In the process of calculating the total cost, a collaborative health impact factor is introduced to dynamically adjust the maintenance priority of the current wind turbine.

[0011] Preferably, S2 specifically includes: In the intelligent operation and maintenance scheduling optimization algorithm for wind farms, the total maintenance time requirement is generated based on the maintenance time required for each wind turbine and combined with the collaborative maintenance time saving factor.

[0012] Preferably, S2 specifically includes: Based on the total maintenance time requirement, combined with the number of available maintenance personnel and the maximum daily maintenance time, operational resource constraints are constructed.

[0013] The beneficial effects of the technical solution of the present invention are: 1. By collecting and preprocessing multi-dimensional operation data of wind farms, and combining it with long short-term memory networks to predict the remaining lifespan of each wind turbine, an accurate assessment of the health status of wind turbines is achieved. This provides a data-driven scientific basis for intelligent operation and maintenance scheduling, avoids the blindness of traditional experience-based judgments, and reduces resource waste and failure risks caused by premature or late maintenance.

[0014] 2. By introducing a wind turbine failure risk assessment algorithm, the importance weight of the wind turbine is determined based on the proportion of historical power generation, and a failure risk coefficient is generated by combining the remaining life of the wind turbine. This enables the accurate quantification of the failure risk of each wind turbine, thereby identifying key equipment that has a significant impact on the overall power generation efficiency of the wind farm. This allows for the priority allocation of maintenance resources, optimization of resource allocation efficiency, and reduction of significant economic losses caused by the failure of key wind turbines.

[0015] 3. Design collaborative health impact factors. By comprehensively considering the health status and operational correlation of adjacent wind turbines, dynamically adjust the maintenance priority of each wind turbine, effectively identify high-risk wind turbines that may cause cascading failures, and prioritize their maintenance to enhance the overall stability of the wind farm operation and reduce the possibility of systemic risks caused by a single wind turbine failure.

[0016] 4. By introducing a collaborative maintenance time-saving factor, the resource-sharing effect when adjacent wind turbines are maintained simultaneously is quantified, which effectively reduces the total maintenance time requirement, improves the utilization efficiency of maintenance resources, shortens the maintenance cycle, reduces downtime caused by maintenance, further reduces power generation loss, and improves the overall economic benefits of the wind farm. Attached Figure Description

[0017] Figure 1 This is a flowchart of a wind farm intelligent operation and maintenance scheduling method based on equipment remaining life prediction, as described in this invention. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Unless otherwise defined, 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 invention pertains.

[0020] The following description, in conjunction with the accompanying drawings, details a specific scheme for an intelligent operation and maintenance scheduling method for wind farms based on equipment remaining life prediction provided by this invention.

[0021] See attached document Figure 1 The diagram illustrates a flowchart of a wind farm intelligent operation and maintenance scheduling method based on equipment remaining life prediction, according to an embodiment of the present invention. The method includes the following steps: S1. Collect and preprocess multidimensional operating data, and predict the remaining life of the wind turbine based on the preprocessed multidimensional operating data; based on the remaining life of the wind turbine, generate the failure risk coefficient of the wind turbine through the wind turbine failure risk assessment algorithm.

[0022] Multidimensional operational data is collected from wind farm sensors and recording systems, such as historical databases of wind farms, covering power output, temperature, fault history (number and type of faults), historical power generation, etc.; the multidimensional operational data is preprocessed, such as by Z-Score standardization, to obtain preprocessed multidimensional operational data; Based on preprocessed multidimensional operational data, a Long Short-Term Memory (LSTM) network is used to predict the remaining lifespan of each wind turbine, forecasting the operating time before a future failure and outputting the remaining lifespan of each turbine. ; The core of the wind turbine failure risk assessment algorithm lies in determining the importance weight of each turbine by utilizing its historical power generation ratio, and then generating a failure risk coefficient for each turbine based on its remaining lifespan. The calculation formula is as follows: , in, Indicates the first The failure risk factor of typhoon generators; Indicates the first The remaining lifespan of the typhoon generator; Indicates the first The importance weight of typhoon generators is calculated using the following formula: ,in, Indicates the first The historical power generation of the typhoon turbines serves as the basis for weighting calculations, reflecting the actual power generation contribution of the turbines. This data is obtained from the historical database of wind farms. Indicates ownership of the wind farm The total historical power generation of the typhoon generators is used as the normalized denominator to ensure that the sum of the weights is 1; This represents the adjustment parameter used to control the impact of the remaining lifespan of the fan on the failure risk coefficient. It is determined by expert experience and has a value range of [0.1, 10]. This represents a very small positive number, used to avoid the denominator being zero. It is determined through expert experience and its typical value is 0.01. The significance of the importance weight of wind turbines is the proportion of their contribution to the power generation benefits of wind farms, while the remaining life reflects the health status of wind turbines. The combination of the two forms the failure risk coefficient of each wind turbine. By quantifying the failure risk of each wind turbine, the equipment with the greatest impact on the overall operation of the wind farm can be identified, providing a scientific basis for intelligent operation and maintenance scheduling.

[0023] S2. Based on the failure risk coefficient of the wind turbine, the optimal operation and maintenance scheduling scheme is generated through the wind farm intelligent operation and maintenance scheduling optimization algorithm.

[0024] Based on the failure risk coefficient of each wind turbine, the optimal operation and maintenance scheduling scheme is generated by the wind farm intelligent operation and maintenance scheduling optimization algorithm with the goal of minimizing the total cost and satisfying the operation and maintenance resource constraints.

[0025] The maintenance decision for each wind turbine is represented by a binary decision variable, where 1 indicates that maintenance is scheduled and 0 indicates that maintenance is not scheduled.

[0026] Total cost is calculated by summing the maintenance costs of all wind turbines and the adjusted downtime losses. Maintenance costs reflect the economic expenditures of performing maintenance, such as labor, material and equipment costs. Downtime losses reflect the economic impact of power generation loss during maintenance. Downtime losses are related to the failure risk coefficient and the synergistic health impact factor. Specifically, the downtime losses are multiplied by the result of dividing the failure risk coefficient by the synergistic health impact factor, so that the weight of downtime losses is dynamically adjusted. The collaborative health impact factor dynamically adjusts the maintenance priority of the current wind turbine based on the remaining lifespan and operational correlation of adjacent wind turbines. For each wind turbine, the set of adjacent wind turbines is traversed, and the ratio of the remaining lifespan of adjacent wind turbines to the remaining lifespan of the current wind turbine is calculated by combining the correlation strength in the adjacency matrix. The collaborative health impact factor is obtained by weighted summation. The larger the collaborative health impact factor, the better the health status of adjacent wind turbines, and the lower the maintenance priority of the current wind turbine. Conversely, the smaller the collaborative health impact factor, the worse the health status of adjacent wind turbines, and the higher the maintenance priority of the current wind turbine, so as to effectively avoid cascading failures. To avoid division by zero errors in the calculation, a small positive number is added to the denominator to ensure numerical stability. The calculation formula is as follows: , in, The total cost is represented by , which includes maintenance costs and downtime losses. Minimizing the total cost is taken as the objective function of optimization. Minimizing the total cost means finding the optimal operation and maintenance scheduling scheme. Indicates all Typhoon machines sum; It is a binary decision variable, representing the first... Whether maintenance is scheduled for the typhoon generator: 1 indicates maintenance is scheduled, 0 indicates maintenance is not scheduled. Indicates the first The maintenance costs of typhoon turbines can be obtained through wind farm operation and maintenance records or data provided by equipment suppliers, including labor, material, and equipment costs. Indicates the first The downtime loss of typhoon turbines, i.e. the cost of lost power generation due to maintenance or failure, is calculated based on downtime, power generation loss, and electricity price. This represents the collaborative health impact factor, used to reflect the influence of the health status of adjacent wind turbines on the current wind turbine maintenance priority; The weight represents the synergistic influence, used to adjust the degree of influence of the health status of adjacent wind turbines on the maintenance priority of the current wind turbine. It is determined by expert experience and has a value range of [0.1, 0.5]. Indicates the first The adjacent wind turbines of the typhoon turbine, and ; It is a collection of adjacent wind turbines; Represents the elements of the adjacency matrix, reflecting the first... Typhoon machine and the The correlation strength of typhoon turbines is used to quantify the operational correlation between turbines. It is calculated based on the wind farm layout, such as the reciprocal of the distance, and the value range is [0,1], where 0 indicates no correlation and 1 indicates strong correlation. Indicates the first The remaining lifespan of the typhoon generator; This indicates the adjusted downtime losses.

[0027] Meanwhile, considering the constraints of operation and maintenance resources, the total maintenance time is ensured to not exceed the available resource limit. The total maintenance time requirement is based on the maintenance time required for each wind turbine and is adjusted in combination with the collaborative maintenance time saving factor. When adjacent wind turbines are maintained at the same time, the total maintenance time will be reduced due to resource sharing, such as sharing maintenance equipment or personnel.

[0028] The formula for total maintenance time requirement is expressed as follows: , in, Indicates the total maintenance time requirement; Indicates the first The time required for typhoon turbine maintenance is obtained by statistically analyzing historical operation and maintenance data in the wind farm's historical database and equipment manufacturer's standard manuals. This represents the collaborative maintenance time saving factor, reflecting the resource sharing effect of simultaneous maintenance of adjacent wind turbines; It is a binary decision variable, representing the first... Whether maintenance is scheduled for the typhoon generator: 1 indicates maintenance is scheduled, 0 indicates maintenance is not scheduled. This represents the collaborative maintenance time saving coefficient, used to control the degree of resource saving in collaborative maintenance. Its value ranges from [0,1] and is determined through expert experience. Indicates the number of available maintenance personnel; This indicates the maximum daily maintenance time.

[0029] Genetic algorithms are used for iterative optimization, continuously adjusting decision variables and prioritizing solutions with low total cost that meet operational resource constraints, until the optimal operational scheduling solution is converged, ensuring both economy and feasibility.

[0030] The specific implementation of the genetic algorithm is as follows: First, an initial population is constructed, where each individual represents an operation and maintenance scheduling scheme, i.e., a combination of binary decision variables. Then, the fitness of each individual is calculated; the fitness function is determined by the optimization objective function based on total cost. Individuals with lower total cost and satisfying operation and maintenance resource constraints have higher fitness. Next, high-fitness individuals are retained through selection, and new solutions, i.e., new operation and maintenance scheduling schemes, are generated through crossover. Random perturbations are then introduced through mutation to increase solution diversity and avoid getting trapped in local optima. As iterations proceed, the population gradually evolves, low-fitness individuals are eliminated, and high-fitness individuals accumulate. Finally, after multiple iterations, if the change in the objective function is very small or the fitness of the population does not significantly improve over several generations, it is considered that the optimal or near-optimal operation and maintenance scheduling scheme has been converged, achieving cost minimization and rational resource allocation.

[0031] The optimized objective function is expressed as follows: , in, This represents minimizing the total cost; Indicates constraints; Indicates operational resource constraints; This represents a binary decision variable constraint.

[0032] By comprehensively considering maintenance costs and downtime losses, economic expenditures are minimized and operation and maintenance costs are reduced; through the resource-sharing effect of collaborative maintenance, maintenance time requirements are effectively reduced and resource utilization efficiency is improved.

[0033] All parameters or data involved in the above-mentioned intelligent operation and maintenance scheduling method for wind farms based on equipment remaining life prediction have been standardized using Z-Score.

[0034] The order of the embodiments is for illustrative purposes only and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0035] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0036] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for intelligent operation and maintenance scheduling of wind farms based on equipment remaining life prediction, characterized in that, Includes the following steps: S1. Collect and preprocess multidimensional operating data, and predict the remaining life of the wind turbine based on the preprocessed multidimensional operating data; based on the remaining life of the wind turbine, generate the failure risk coefficient of the wind turbine through the wind turbine failure risk assessment algorithm; S2. Based on the failure risk coefficient of the wind turbine, the optimal operation and maintenance scheduling scheme is generated through the wind farm intelligent operation and maintenance scheduling optimization algorithm.

2. The intelligent operation and maintenance scheduling method for wind farms based on equipment remaining life prediction according to claim 1, characterized in that, S1 specifically includes: Long Short-Term Memory (LSTM) networks are used to predict the operating time of wind turbines before they fail in the future, and the remaining lifespan of each wind turbine is output.

3. The intelligent operation and maintenance scheduling method for wind farms based on equipment remaining life prediction according to claim 2, characterized in that, S1 specifically includes: In the implementation of the wind turbine failure risk assessment algorithm, the importance weight of the wind turbine is calculated based on its historical power generation, and the failure risk coefficient of each wind turbine is generated by combining the remaining life of each wind turbine.

4. The intelligent operation and maintenance scheduling method for wind farms based on equipment remaining life prediction according to claim 1, characterized in that, S2 specifically includes: The intelligent operation and maintenance scheduling optimization algorithm for wind farms aims to minimize total cost while satisfying operation and maintenance resource constraints, and generates the optimal operation and maintenance scheduling scheme.

5. The intelligent operation and maintenance scheduling method for wind farms based on equipment remaining life prediction according to claim 4, characterized in that, S2 specifically includes: The total cost is obtained by summing the maintenance costs of all wind turbines and the adjusted downtime losses.

6. The intelligent operation and maintenance scheduling method for wind farms based on equipment remaining life prediction according to claim 5, characterized in that, S2 specifically includes: In the process of calculating the total cost, a collaborative health impact factor is introduced to dynamically adjust the maintenance priority of the current wind turbine.

7. The intelligent operation and maintenance scheduling method for wind farms based on equipment remaining life prediction according to claim 4, characterized in that, S2 specifically includes: In the intelligent operation and maintenance scheduling optimization algorithm for wind farms, the total maintenance time requirement is generated based on the maintenance time required for each wind turbine and combined with the collaborative maintenance time saving factor.

8. The intelligent operation and maintenance scheduling method for wind farms based on equipment remaining life prediction according to claim 7, characterized in that, S2 specifically includes: Based on the total maintenance time requirement, combined with the number of available maintenance personnel and the maximum daily maintenance time, operational resource constraints are constructed.

Images