Wind power plant operation and maintenance planning method

Abstract

The invention relates to the technical field of wind power generation, in particular to a wind power plant operation and maintenance planning method which comprises the following steps: acquiring and preprocessing historical operation data of a wind power plant; extracting position distribution of fans in the wind power plant, constructing a fault prediction model, and outputting a spare part demand prediction result; constructing a spare part scheduling model, and outputting a spare part scheduling plan; and constructing an operation and maintenance personnel scheduling model based on the spare part scheduling plan, and outputting an operation and maintenance personnel scheduling result. The invention aims to solve the technical problem of high-efficiency and low-cost operation and maintenance of the small-capacity wind generating set.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power generation, and in particular to a wind farm operation and maintenance planning method. Background Art

[0002] In the context of the current global energy transformation, wind energy, as one of the important renewable energy sources, is undergoing unprecedented changes and upgrades in its power generation technology and application. In particular, small wind turbines of 1.5MW and 2.0MW have played a key role in the wind power market in the past decade, but as these devices gradually exceed the manufacturer's warranty period, they face more severe operation and maintenance challenges. This type of early wind turbine may have deficiencies in the design and production stage, such as improper material selection, unreasonable structural design, or lax process control, resulting in poor equipment operation stability and high failure rate, thus affecting the overall power generation efficiency and economic benefits. As time goes by, after the wind turbine is out of warranty, the original manufacturer's support weakens, the spare parts procurement cycle is long and the cost is high, and the variety of spare parts due to the diversity of models brings huge pressure on inventory management and cost control. In particular, the shortage of spare parts directly leads to the extension of unplanned downtime of wind turbines, affecting power generation and operation and maintenance efficiency.

[0003] At present, as large power generation groups have put forward higher requirements for real-time monitoring and operation and maintenance response of wind turbines, although the wind farm control center can collect a large amount of wind turbine operation data, the traditional operation and maintenance management model is difficult to respond quickly to such small wind turbines, which ultimately leads to a significant increase in operation and maintenance costs over the years. The traditional operation and maintenance model has been unable to adapt to the high-efficiency and low-cost operation and maintenance needs of modern wind power generation systems. Therefore, it is urgent to develop a set of operation and maintenance planning methods for wind farms. Summary of the invention

[0004] In order to solve the technical problem of efficient and low-cost operation and maintenance of small-capacity wind turbines, the present invention aims to provide a wind farm operation and maintenance planning method, and the technical solution adopted is as follows:

[0005] A wind farm operation and maintenance planning method, the method comprising:

[0006] Obtain historical operation data of wind farms and perform preprocessing;

[0007] Extract the location distribution of wind turbines in the wind farm, build a fault prediction model, and output the spare parts demand prediction results;

[0008] Build spare parts scheduling model and output spare parts scheduling plan;

[0009] Build an operation and maintenance personnel scheduling model based on the spare parts scheduling plan and output the operation and maintenance personnel scheduling results.

[0010] Furthermore, the historical operation data of the wind farm includes at least wind turbine operation status data, historical fault data, spare parts consumption data, wind turbine location data and wind turbine warranty period data.

[0011] Furthermore, the location distribution of wind turbines in the wind farm is extracted, a fault prediction model is constructed, and the output of spare parts demand prediction results includes:

[0012] Integrate wind farm turbine location data, warehouse location data, topographic data, and road network data;

[0013] Predict wind turbine failures in wind farms based on historical operation data of wind farms and predict spare parts demand in the future;

[0014] Calculate the initial cost of a failed wind turbine based on spare parts requirements.

[0015] Furthermore, based on the historical operation data of the wind farm, the prediction of wind turbine failures in the wind farm and the prediction of spare parts demand in the future also include:

[0016] Obtain the location data of the faulty fans, calculate the straight-line distance between the faulty fans, and divide the faulty fans into clusters;

[0017] Calculate the cluster effect coefficient and isolation coefficient of the faulty fan;

[0018] Predict the demand for spare parts in the future based on the cluster effect coefficient and isolation coefficient.

[0019] Furthermore, the initial cost of a failed wind turbine based on the spare parts demand is calculated as follows:

[0020] Based on the initial cost of spare parts demand, Monte Carlo simulation is used to simulate the uncertainty of spare parts demand and obtain the probability distribution of spare parts demand;

[0021] Monte Carlo simulation is used to optimize the initial cost based on the spare parts cost within and outside the warranty period, taking into account the wind turbine operating years and warranty period.

[0022] Furthermore, based on the historical operation data of the wind farm, the prediction of wind turbine failures in the wind farm and the prediction of spare parts demand in the future also include:

[0023] Using clustering algorithm, wind turbines are divided into different clusters according to preset distances;

[0024] Calculate the cluster effect coefficient and isolation coefficient based on the average fan density of the fans in the cluster;

[0025] Identify isolated fans based on the isolation coefficient and set up temporary spare parts storage points for the isolated fans.

[0026] Furthermore, a spare parts scheduling model is constructed, and the output spare parts scheduling plan includes:

[0027] Obtain the fluctuation distribution of spare parts supply;

[0028] Define the number of decision variables;

[0029] With the goal of minimizing the spare parts replacement cost, a spare parts scheduling model is constructed. The spare parts demand forecast results considering the influence of wind turbine location and warranty period are input to solve the optimal spare parts scheduling plan.

[0030] Furthermore, the spare parts scheduling model includes:

[0031] The minimum spare parts inventory level constraint is defined as the spare parts inventory level not falling below the preset minimum spare parts inventory level at any time;

[0032] The fan downtime limit constraint is defined as any fan downtime not less than the preset maximum fan downtime;

[0033] The spare parts inventory level constraint is defined as the spare parts inventory level at any time that can cover the spare parts demand adjusted by the clustering effect coefficient and the isolation coefficient;

[0034] The warranty period constraint is defined as the spare parts inventory level at any time that can cover the spare parts demand adjusted by the warranty period weight factor.

[0035] Furthermore, an operation and maintenance personnel scheduling model is constructed based on the spare parts scheduling plan, and the output of the operation and maintenance personnel scheduling results includes:

[0036] Define decision variables for operations personnel;

[0037] Enter the preset operation and maintenance system;

[0038] With the goal of minimizing the cost of operation and maintenance personnel, a scheduling model for operation and maintenance personnel is constructed to solve the optimal scheduling plan.

[0039] Furthermore, the operation and maintenance personnel scheduling model includes:

[0040] Hour constraint, defined as the total number of hours worked by the operation and maintenance personnel at any time not exceeding the preset maximum allowed working hours;

[0041] The first number constraint is defined as the minimum number of operation and maintenance personnel required to complete the spare parts replacement task at any time;

[0042] The second headcount constraint is defined as the number of O&M personnel at any time that meets the spare parts demand adjusted by the cluster effect coefficient and the isolated wind turbine mark;

[0043] The third number constraint is defined as the number of operation and maintenance personnel at any time that meets the spare parts demand adjusted by the warranty period weight coefficient.

[0044] The present invention has the following beneficial effects:

[0045] The present invention provides a wind farm operation and maintenance planning method, which predicts the failure probability of wind turbines and, based on the predicted demand for spare parts, further considers the impact of wind turbine locations and warranty periods on spare parts replacement time and cost, so as to help decision makers understand the fluctuation range of spare parts demand, avoid excessive reserve waste, and avoid the problem of wind turbine downtime and high procurement costs caused by insufficient spare parts. Ultimately, the optimal operation and maintenance personnel scheduling plan and spare parts dispatching plan are extracted to reduce personnel and spare parts costs and ensure that spare parts are in place in a timely manner. The model can be adjusted according to the specific conditions of the wind farm to adapt to wind farms of different sizes and distributions. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or 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. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 A schematic flow chart of a wind farm operation and maintenance planning method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a wind farm operation and maintenance planning method proposed by the present invention, its specific implementation method, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0049] 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 invention belongs.

[0050] A specific scheme of a wind farm operation and maintenance planning method provided by the present invention is described in detail below with reference to the accompanying drawings.

[0051] See also Figure 1 , which shows a method flow chart of a wind farm operation and maintenance planning method provided by an embodiment of the present invention, a wind farm operation and maintenance planning method, the method comprising:

[0052] Step S100: Obtain historical operation data of the wind farm and perform preprocessing; the historical operation data of the wind farm at least includes wind turbine operation status data, historical fault data, spare parts consumption data, wind turbine location data and wind turbine warranty period data; the wind turbine operation status data at least includes start, stop and fault status; the historical fault data at least includes: fault type, occurrence time, and maintenance time; the spare parts consumption data at least includes: spare part name, consumption time, and consumption quantity; the wind turbine location data at least includes: latitude and longitude coordinates, altitude; the preprocessing of these data includes but is not limited to removing outliers, filling missing values ​​and standardizing values;

[0053] Step S200: extracting the location distribution of wind turbines in the wind farm, building a fault prediction model, and outputting spare parts demand prediction results; specifically, the specific location information of the wind turbines can be extracted through a geographic information system (GIS), and a machine learning or deep learning model can be trained in combination with historical fault data to predict the types of faults that may occur in the future and their probabilities, and then the number of spare parts required can be determined by predicting possible faults in advance;

[0054] Step S200 specifically includes:

[0055] Step S210: Integrate wind turbine location data, warehouse location data, topographic data and road network data in the wind farm; specifically, collect the exact geographical location coordinates of each wind turbine in the wind farm through the GIS system; the warehouse location data is the location coordinates of the spare parts warehouse, which can be used to calculate the shortest path from the warehouse to the wind turbine; the topographic data at least includes slope, soil type, etc., which can be obtained through remote sensing images or topographic maps; the road network data at least includes road type, width, length, etc., which helps to determine the best route for spare parts transportation; integrate the above data together to provide comprehensive information support for subsequent fault prediction and spare parts demand prediction;

[0056] Step S220: predicting wind turbine failures in the wind farm based on the historical operation data of the wind farm, and predicting the demand for spare parts and wind turbine downtime in the future; specifically, extracting features including but not limited to average wind speed, temperature change, generator vibration frequency, operating hours, wind turbine load, ambient humidity, and power output from the historical operation data of the wind farm, and using, for example, random forest, support vector machine, neural network model to perform the prediction task, and the expression is:

[0057] P f (t) = f(x t ;θ)

[0058] Where P f (t) is the probability of fan failure at time t; x trepresents the feature vector input at time t; θ represents the first model parameter; f(.) represents the prediction model;

[0059] Step S220 specifically includes:

[0060] Step S221: Obtain the location data of the faulty wind turbine, calculate the straight-line distance between the faulty wind turbines and divide the faulty wind turbines into clusters, the expression of which is:

[0061]

[0062] Where, d im represents the straight-line distance from faulty fan i to faulty fan m; (x i -y i ) 2 and (x m -y m ) 2 are the coordinates of faulty fan i and faulty fan m respectively;

[0063] Then, the wind turbines are divided into different clusters according to the preset distance threshold r. Clustering algorithms (such as K-means) or distance-based clustering methods (such as DBSCAN) can be used to achieve this goal. The expression is:

[0064]

[0065] In the formula, C k represents the mth cluster;

[0066] Step S222: Calculate the cluster effect coefficient and isolation coefficient of the faulty wind turbine, and the expression is:

[0067]

[0068] In the formula, ρ i represents the average fan density of fan i; A i represents the cluster area of ​​wind turbine i; r is the preset distance threshold, which is used to define the wind turbine with scarf failure; d im represents the straight-line distance from faulty fan i to faulty fan m;

[0069]

[0070] In the formula, C i represents the cluster effect coefficient of wind turbine i; max m (ρ m ) represents the maximum average fan density among all fans;

[0071]

[0072] In the formula, Ii Indicates the isolation flag of wind turbine i; τ indicates the preset isolation prefabrication, if ρ i <τ, the wind turbine is considered to be an isolated wind turbine; by grouping wind turbines with similar locations into the same cluster, the impact of location factors on spare parts demand can be more effectively considered. Wind turbines with similar locations may share some spare parts, which is conducive to the centralized storage and scheduling of spare parts; considering isolated wind turbines: for isolated wind turbines, you can consider replenishing spare parts in advance. Because these wind turbines may find it difficult to quickly obtain spare parts support, you can consider setting up temporary spare parts storage points for isolated wind turbines to reduce the transportation distance and time from the spare parts warehouse to the isolated wind turbines.

[0073] Step S223: predicting the demand for spare parts in the future based on the cluster effect coefficient and the isolation coefficient, the expression is:

[0074]

[0075] Where N(t) represents the spare parts demand at time t; P f (t) represents the probability of fan failure at time t; C i Represents the cluster effect coefficient of the fan; I i represents the fan isolation coefficient; η i represents the spare parts consumption coefficient of the wind turbine, reflecting the number of spare parts consumed by the wind turbine when a failure occurs; N represents the total number of wind turbines in the wind farm; specifically, when constructing the spare parts demand prediction model, the cluster effect coefficient and the isolated wind turbine mark are introduced to reflect the different characteristics of wind turbines in the cluster and isolated wind turbines; for example, for wind turbines in the cluster, the spare parts reserve can be appropriately increased because they face similar operating environments and higher failure probabilities; for isolated wind turbines, it is possible to consider replenishing spare parts in advance because these wind turbines may find it difficult to quickly obtain spare parts support; by grouping wind turbines with similar locations into the same cluster, the impact of location factors on spare parts demand can be more effectively considered, and wind turbines with similar locations may share certain spare parts, which is conducive to the centralized storage and scheduling of spare parts; at the same time, when constructing the spare parts transportation time model and cost model, the wind turbines in the cluster can be given priority, because the wind turbines in the cluster are usually geographically close and can share a spare parts transportation route, thereby reducing transportation time and cost. It should be noted that isolated wind turbines may require separate spare parts transportation routes due to their remote geographical location. Therefore, the particularity of these wind turbines needs to be considered when building the model. It can be considered to set up temporary spare parts storage points for isolated wind turbines to reduce the transportation distance and time from the spare parts warehouse to the isolated wind turbines.

[0076] Step S230: Calculating the initial cost of the faulty wind turbine based on the spare parts demand;

[0077] Step S230 specifically includes:

[0078] Step S231: Based on the initial cost of the spare parts demand, the uncertainty of the spare parts demand is simulated by Monte Carlo simulation to obtain the probability distribution of the spare parts demand, which is expressed as:

[0079] N(t)~D(gP f (t),C i ,I i ,η i ;θ′)

[0080] Where N(t) is the probability distribution of spare parts demand at time t; D(.) represents the probability distribution function obtained by Monte Carlo simulation; θ′ represents the second model parameter;

[0081] Step S232: using Monte Carlo simulation, combined with the wind turbine operating years and warranty period, the initial cost is optimized based on the spare parts cost within and outside the warranty period, and the expression is:

[0082]

[0083] In the formula, C total represents the initial cost; E[N(t)] represents the expected value of the spare parts demand at time t; C piece Indicates the cost of a single spare part; represents the weight coefficient of the warranty period. If the fan i is still within the warranty period within time t, it is w in , otherwise w out ; It can be defined as:

[0084]

[0085] Where, T warranty Indicates the end time of the fan warranty period.

[0086] In summary, the demand for spare parts in the future is predicted based on the cluster effect coefficient and the isolated wind turbine mark, and then based on the initial cost of the spare parts demand, the uncertainty of the spare parts demand is simulated by Monte Carlo simulation to obtain the probability distribution of the spare parts demand. Finally, the Monte Carlo simulation is used to optimize the initial cost based on the spare parts cost within and outside the warranty period in combination with the wind turbine operating years and warranty period. This method quantifies the impact of wind turbine location and warranty period on the spare parts demand forecast results, improves the accuracy of spare parts management and cost control, and can help wind farm managers better understand the uncertainty of spare parts demand and formulate a more scientific and reasonable spare parts reserve strategy, thereby reducing operating costs and improving operation and maintenance efficiency.

[0087] Step S300: construct a spare parts scheduling model and output a spare parts scheduling plan;

[0088] Step S300 specifically includes:

[0089] Step S310: Obtain the fluctuation distribution of spare parts supply based on historical data and information provided by suppliers, and its expression is:

[0090] S(t)~D supply

[0091] In the formula, S(t) represents the spare parts supply at time t, D supply Represents the fluctuation distribution of spare parts supply;

[0092] Step S320: define a spare quantity decision variable, denoted as Q(t), which is the quantity of spare parts to be purchased at time t;

[0093] Step S330: With the goal of minimizing the spare parts replacement cost, a spare parts scheduling model is constructed, and the spare parts demand forecast results considering the influence of the wind turbine location and warranty period are input to solve the optimal spare parts scheduling solution, which is expressed as:

[0094]

[0095] c re (t) represents the spare parts replacement cost at time t;

[0096] The minimum spare parts inventory level constraint is defined as the spare parts inventory level not being lower than the preset minimum spare parts inventory level at any time. Its expression is:

[0097]

[0098] Where t′ represents any time point before time t; N(t′) represents the spare parts demand calculated in step S200; L represents the minimum spare parts inventory level; By comparing the cumulative number of spare parts purchased from the start time point to the current time point t Subtract the total spare parts demand during this period This ensures that there is always enough spare parts in stock to deal with fan failures

[0099] The fan downtime limit constraint is defined as any fan downtime not less than the preset maximum fan downtime, and its expression is:

[0100]

[0101] Where, T down Indicates the fan downtime; D max Represents the maximum fan downtime; by accumulating the fan downtime T from the start time point to the current time point t down (t′) and the maximum allowable downtime D max This is done by making comparisons, which helps ensure that the normal operation of the wind farm is not affected by long-term outages;

[0102] The spare parts inventory level constraint is defined as the spare parts inventory level at any time that can cover the spare parts demand adjusted by the cluster effect coefficient and the isolated wind turbine flag. Its expression is:

[0103]

[0104] Where P f (t) represents the probability of fan failure at time t; C i Represents the cluster effect coefficient of the fan; I i represents the fan isolation coefficient; η i represents the spare parts consumption coefficient of the wind turbine, reflecting the number of spare parts consumed by the wind turbine when a failure occurs; N represents the total number of wind turbines in the wind farm; by comparing the cumulative number of spare parts purchased from the start time point to the current time point t Adjusted spare parts demand This can be achieved by taking into account the different characteristics of wind turbines in the cluster and isolated wind turbines to better meet actual needs;

[0105] The warranty period constraint is defined as the spare parts inventory level at any time that can cover the spare parts demand adjusted by the warranty period weight coefficient. Its expression is:

[0106]

[0107] In the formula, Represents the warranty period weight coefficient; it compares the cumulative number of spare parts purchased from the start time point to the current time point tt Adjusted spare parts demand To achieve this, we take into account the low maintenance cost of the wind turbine during the warranty period to reduce the spare parts cost expenditure; then, we use linear programming or mixed integer programming to solve the optimal spare parts scheduling solution. In the model, the spare parts demand N(t) is used as input, and the constraints are used to ensure that the spare parts inventory level meets the spare parts demand in the future; through the cluster effect coefficient C i , we can see that the spare parts demand of wind turbines in the cluster may be higher because they face similar operating environments and higher failure probabilities. Therefore, in the model, we can adjust the spare parts demand to reflect this effect. On the other hand, isolated wind turbines may need more spare parts reserves to cope with potential delays due to their remote locations. In the model, we can also adjust the spare parts demand to reflect this effect. Since the maintenance cost of wind turbines within the warranty period is lower, while the maintenance cost of wind turbines outside the warranty period is higher, in the model, we use the warranty period weight coefficient To adjust the spare parts demand and ensure that the spare parts cost expenditure is minimized during the warranty period; by constructing a spare parts scheduling model, this embodiment can more effectively manage and schedule spare parts, reduce unnecessary inventory costs, and ensure that there are enough spare parts to deal with wind turbine failures; by considering the spare parts demand, cluster effect coefficient, isolated wind turbine mark and the impact of the warranty period, the spare parts demand can be more accurately predicted and the spare parts cost can be optimized.

[0108] Step S400: constructing an operation and maintenance personnel scheduling model based on the spare parts scheduling plan, and outputting the operation and maintenance personnel scheduling results;

[0109] Step S400 specifically includes: defining decision variables for operation and maintenance personnel;

[0110] Step S410: input the preset operation and maintenance system;

[0111] Step S420: With the goal of minimizing the cost of operation and maintenance personnel, an operation and maintenance personnel scheduling model is constructed to solve the optimal scheduling plan; the operation and maintenance personnel scheduling model includes: an hour constraint, which is defined as the total working hours of the operation and maintenance personnel at any time not exceeding the preset maximum allowed working hours; a first number constraint, which is defined as the number of operation and maintenance personnel at any time at least reaching the minimum number of personnel required to complete the spare parts replacement task; a second number constraint, which is defined as the number of operation and maintenance personnel at any time meeting the spare parts demand adjusted by the cluster effect coefficient and the isolated wind turbine mark; a third number constraint, which is defined as the number of operation and maintenance personnel at any time meeting the spare parts demand adjusted by the warranty period weight coefficient; specifically, the constraints of the operation and maintenance personnel scheduling model are similar to those of the spare parts scheduling model, so they will not be repeated again; in step S400, also considering the spare parts demand N(t), we can deduce the number of spare parts replacement tasks that need to be completed at each time point, which will directly determine the number of operation and maintenance personnel that need to be arranged at a specific time point; through the cluster effect coefficient C i , we can understand that the demand for operation and maintenance personnel for fans in the cluster may be higher because they face similar operating environments and higher failure probabilities. Therefore, in this model, we can reflect this effect through the number of operation and maintenance personnel; by considering the isolated fan sign, isolated fans may require more operation and maintenance personnel to handle them due to their remote geographical location to ensure rapid response. In the model, we can also reflect this effect by adjusting the number of operation and maintenance personnel; the maintenance cost of fans within the warranty period is lower, while the maintenance cost of fans outside the warranty period is higher. Therefore, in this model, we adjust the number of operation and maintenance personnel through the warranty period weight coefficient to ensure that the cost expenditure of operation and maintenance personnel is minimized during the warranty period; Operation and maintenance personnel scheduling: by constructing an operation and maintenance personnel scheduling model, we can reasonably arrange the working hours of operation and maintenance personnel, improve work efficiency, and ensure the timeliness and effectiveness of fan maintenance work.

[0112] In summary, this wind farm operation and maintenance planning method aims to improve the operation and maintenance efficiency and cost-effectiveness of wind farms, and realizes refined management of wind farms through a series of steps. On the basis of predicting wind turbine failures, the cluster effect coefficient, isolated wind turbine mark and warranty period are comprehensively considered to build a spare parts demand prediction model. Optimizing spare parts demand prediction helps to reduce spare parts storage costs and reduce delayed maintenance time caused by insufficient spare parts, thereby improving the availability and reliability of wind farms; by building a spare parts scheduling model, efficient scheduling of spare parts is achieved to ensure that spare parts are in place in time and wind turbine downtime is reduced. Improving spare parts scheduling efficiency can reduce wind turbine downtime, increase wind farm power generation, and thus increase economic benefits; by building an operation and maintenance personnel scheduling model, the working hours of operation and maintenance personnel can be reasonably arranged to ensure that there are enough people to handle spare parts replacement and other operation and maintenance tasks at the right time. By integrating the above steps, a complete set of wind farm operation and maintenance planning schemes can be formed to achieve comprehensive optimization of wind farm operation and maintenance strategies, organically combine fault prediction, spare parts demand prediction, spare parts scheduling, and operation and maintenance personnel scheduling to form a closed-loop operation and maintenance management system. Comprehensive optimization of wind farm operation and maintenance strategies helps to improve the overall operational efficiency of wind farms, reduce operation and maintenance costs, and improve the competitiveness of wind farms. This wind farm operation and maintenance planning method achieves the optimization of wind farm operation and maintenance strategies through a series of scientific and reasonable steps and technical means, which is specifically reflected in improving fault prediction accuracy, optimizing spare parts demand prediction, improving spare parts scheduling efficiency, reasonably arranging operation and maintenance personnel scheduling, and comprehensively optimizing wind farm operation and maintenance strategies. These technical effects not only help to improve the operation and maintenance efficiency of wind farms, but also can significantly reduce costs, and ultimately achieve a dual improvement in economic and social benefits.

[0113] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0114] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A wind farm operation and maintenance planning method, characterized in that: The method comprises: Obtain historical operation data of wind farms and perform preprocessing; Extract the location distribution of wind turbines in the wind farm, build a fault prediction model, and output the spare parts demand prediction results; Build spare parts scheduling model and output spare parts scheduling plan; Build an operation and maintenance personnel scheduling model based on the spare parts scheduling plan and output the operation and maintenance personnel scheduling results.

2. A wind farm operation and maintenance planning method according to claim 1, characterized in that: The historical operation data of the wind farm includes at least wind turbine operation status data, historical fault data, spare parts consumption data, wind turbine location data and wind turbine warranty period data.

3. A wind farm operation and maintenance planning method according to claim 1, characterized in that: Extract the location distribution of wind turbines in the wind farm, build a fault prediction model, and output the spare parts demand prediction results including: Integrate wind farm turbine location data, warehouse location data, topographic data, and road network data; Predict wind turbine failures in wind farms based on historical operation data of wind farms and predict spare parts demand in the future; Calculate the initial cost of a failed wind turbine based on spare parts requirements.

4. A wind farm operation and maintenance planning method as claimed in claim 3, characterized in that: Predict wind turbine failures in wind farms based on historical operation data of wind farms, and predict spare parts demand in the future period also include: Obtain the location data of the faulty fans, calculate the straight-line distance between the faulty fans, and divide the faulty fans into clusters; Calculate the cluster effect coefficient and isolation coefficient of the faulty fan; Predict the demand for spare parts in the future based on the cluster effect coefficient and isolation coefficient.

5. A wind farm operation and maintenance planning method as claimed in claim 4, characterized in that: Calculating the initial cost of a failed wind turbine based on spare parts requirements includes: Based on the initial cost of spare parts demand, Monte Carlo simulation is used to simulate the uncertainty of spare parts demand and obtain the probability distribution of spare parts demand; Monte Carlo simulation is used to optimize the initial cost based on the spare parts cost within and outside the warranty period, taking into account the wind turbine operating years and warranty period.

6. A wind farm operation and maintenance planning method as claimed in claim 5, characterized in that: Predict wind turbine failures in wind farms based on historical operation data of wind farms, and predict spare parts demand in the future period also include: Using clustering algorithm, wind turbines are divided into different clusters according to preset distances; Calculate the cluster effect coefficient and isolation coefficient based on the average fan density of the fans in the cluster; Identify isolated fans based on the isolation coefficient and set up temporary spare parts storage points for the isolated fans.

7. A wind farm operation and maintenance planning method according to any one of claims 1 to 6, characterized in that: Build a spare parts scheduling model and output spare parts scheduling plans including: Obtain the fluctuation distribution of spare parts supply; Define the number of decision variables; With the goal of minimizing the spare parts replacement cost, a spare parts scheduling model is constructed. The spare parts demand forecast results considering the influence of wind turbine location and warranty period are input to solve the optimal spare parts scheduling plan.

8. A wind farm operation and maintenance planning method according to claim 7, characterized in that: The spare parts scheduling model includes: The minimum spare parts inventory level constraint is defined as the spare parts inventory level not falling below the preset minimum spare parts inventory level at any time; The fan downtime limit constraint is defined as any fan downtime not less than the preset maximum fan downtime; The spare parts inventory level constraint is defined as the spare parts inventory level at any time that can cover the spare parts demand adjusted by the clustering effect coefficient and the isolation coefficient; The warranty period constraint is defined as the spare parts inventory level at any time that can cover the spare parts demand adjusted by the warranty period weight factor.

9. A wind farm operation and maintenance planning method according to claim 8, characterized in that: Build an operation and maintenance personnel scheduling model based on the spare parts scheduling plan, and output the operation and maintenance personnel scheduling results including: Define decision variables for operations personnel; Enter the preset operation and maintenance system; With the goal of minimizing the cost of operation and maintenance personnel, a scheduling model for operation and maintenance personnel is constructed to solve the optimal scheduling plan.

10. A wind farm operation and maintenance planning method according to claim 9, characterized in that: The operation and maintenance personnel scheduling model includes: Hour constraint, defined as the total number of hours worked by the operation and maintenance personnel at any time not exceeding the preset maximum allowed working hours; The first number constraint is defined as the minimum number of operation and maintenance personnel required to complete the spare parts replacement task at any time; The second headcount constraint is defined as the number of O&M personnel at any time that meets the spare parts demand adjusted by the cluster effect coefficient and the isolated wind turbine mark; The third number constraint is defined as the number of operation and maintenance personnel at any time that meets the spare parts demand adjusted by the warranty period weight coefficient.

Images