Cleaning control method, device and equipment of photovoltaic cleaning robot and storage medium
Through the perception module, the pollution shadow performance relationship surface is generated, and the cleaning motion parameters are optimized. The problem of existing photovoltaic cleaning robots failing to effectively consider the impact of photovoltaic pollution and shadows is achieved, and efficient photovoltaic module cleaning and power generation are achieved. Improve efficiency.
Patent Information
- Application Number
- CN202411373345.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-29
- Publication Date
- 2024-11-05
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing photovoltaic cleaning robots fail to effectively consider the impact of multiple photovoltaic pollution parameters and robot shadows on power generation efficiency, resulting in poor cleaning results and affecting photovoltaic power generation efficiency.
Through the perception module, multi-source data is collected, preprocessed and cross-analysed, the polluted shadow performance relationship surface is generated, multi-objective cleaning evaluation is performed, cleaning motion parameters are optimized, cleaning trajectory parameters are generated that minimizes the impact of shadows, and dynamic shadow trajectory and cleaning paths are realized optimization.
It realizes that while ensuring the cleaning effect of photovoltaic modules, it reduces the impact of cleaning on power generation efficiency and improves photovoltaic power generation efficiency.
Smart Images

Figure CN118900087A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of photovoltaic cleaning, and in particular to a cleaning control method, device, equipment and storage medium of a photovoltaic cleaning robot. Background Art
[0002] With the continuous development of photovoltaic power generation technology and the continuous expansion of its application scope, the cleaning and maintenance of photovoltaic modules has become a key link in ensuring power generation efficiency. During the operation of photovoltaic power stations, dust, leaves, bird droppings and other pollutants will inevitably accumulate on the surface of the modules, which will significantly reduce the power generation efficiency of photovoltaic modules. In order to solve this problem, relevant photovoltaic cleaning robots came into being and became an important tool for cleaning and maintenance of large photovoltaic power stations.
[0003] Existing photovoltaic cleaning robots usually use fixed cleaning paths and cleaning parameters, and lack adaptability to real-time environments and component states. Since the degree and type of pollution in different areas may vary, fixed cleaning parameters may lead to insufficient cleaning of some areas, while other areas may be over-cleaned, resulting in water waste and possible component damage; and the traditional cleaning method does not consider the impact of the shadows generated by the cleaning robot itself during its movement on the power generation efficiency of photovoltaic components. Under strong light conditions, this shadow effect may cause significant power generation losses. In addition, the cleaning method also fails to fully consider factors such as light intensity and power demand of the power station. Cleaning operations may be performed during peak power generation periods, thereby affecting the overall power generation efficiency of the power station. That is, the existing cleaning method fails to consider multiple photovoltaic pollution parameters, as well as the impact of robot shadows and photovoltaic power generation status, so that not only does it fail to achieve a good cleaning effect, but it also affects the photovoltaic power generation efficiency. Summary of the invention
[0004] The main purpose of the present invention is to solve the problem that the existing cleaning method fails to take into account various photovoltaic pollution parameters, as well as the influence of robot shadow and photovoltaic power generation status, so that not only a good cleaning effect cannot be achieved, but also the photovoltaic power generation efficiency is affected.
[0005] A first aspect of the present invention provides a cleaning control method for a photovoltaic cleaning robot, the photovoltaic cleaning robot comprising a perception module and a cleaning module, the cleaning control method for the photovoltaic cleaning robot comprising: collecting multi-source perception data of a photovoltaic component to be cleaned through the perception module, and preprocessing the multi-source perception data to obtain component performance status information and component contamination degree information; performing a cross-analysis of multiple power generation influencing parameters on the component performance status information and the component contamination degree information to obtain a contamination shadow performance relationship surface of the robot shadow's influence on component power generation, and performing a multi-objective cleaning evaluation on the contamination shadow performance relationship surface to obtain an initial cleaning strategy that considers the shadow influence; performing an optimization calculation of the cleaning motion parameters of the initial cleaning strategy to obtain a cleaning trajectory parameter that minimizes the shadow influence, and based on the cleaning trajectory parameter, using the cleaning module to perform a cleaning operation on the photovoltaic component to be cleaned to obtain an initial cleaning effect and initial power generation performance data; based on the initial power generation performance data, evaluating the cleaning effect of the initial cleaning result to obtain an initial cleaning evaluation result, and based on the initial cleaning evaluation result, adjusting the cleaning trajectory parameter to generate a final cleaning result of the photovoltaic cleaning robot on the photovoltaic component to be cleaned.
[0006] Optionally, in a first implementation of the first aspect of the present invention, the multi-source perception data includes current and voltage data, light intensity data, component surface image data, robot position data and component real-time power generation data of the photovoltaic component, and the pre-processing of the multi-source perception data to obtain component performance status information and component contamination degree information includes: performing wavelet transform on the current and voltage data to obtain current and voltage data with high-frequency noise removed, and calculating the component output power before cleaning the photovoltaic component based on the current and voltage data, and normalizing the light intensity data to obtain standardized light intensity data, and performing Gaussian blur and edge processing on the component surface image. Edge detection is performed to obtain a pollution distribution image after edge enhancement, and Kalman filtering is performed on the robot position data to obtain position posture data, and a preset simulation power generation model is used to calculate the power generation efficiency of the standardized light intensity data and the component output power to obtain the component power generation efficiency coefficient; timestamp alignment and interpolation are performed on the current and voltage data with high-frequency noise removed, the component output power, the standardized light intensity data, the pollution distribution map, the position posture data and the component power generation efficiency coefficient to obtain time-synchronized component performance status information; the pollution coverage and pollution degree of the pollution distribution image are calculated to obtain component pollution degree information.
[0007] Optionally, in a second implementation of the first aspect of the present invention, the cross-analysis of multiple power generation influencing parameters of the component performance status information and the component pollution degree information is performed to obtain a pollution shadow-performance relationship surface of the robot shadow's impact on component power generation, including: using a preset robot geometric surface to perform position and posture projection transformation on the corresponding position and posture data in the component performance status information to obtain a dynamic shadow contour of the component surface, and based on the preset multiple shadow states and the dynamic shadow contour, analyzing the power generation loss of the component power generation efficiency coefficient in the component performance status information to obtain the power generation loss data of the local shadow, and based on the corresponding current and voltage data in the component performance status information, Determine the changing relationship of the photovoltaic components under different shadow conditions and generate the electrical correlation between components; perform multivariate regression analysis on the component pollution level information and the power generation loss data of the local shadow to generate a preliminary relationship surface, and based on the electrical correlation and the preliminary relationship surface, perform power generation characteristic analysis on the photovoltaic components to obtain a global power generation influence coefficient; use preset historical shadow performance data to perform time series analysis on the global power generation influence coefficient to generate a shadow influence prediction surface that considers time factors, and based on the component pollution level information, fuse the pollution level data with the shadow influence prediction surface to obtain a pollution-shadow-performance relationship surface of the robot shadow's impact on component power generation.
[0008] Optionally, in a third implementation of the first aspect of the present invention, a multi-objective cleaning evaluation is performed on the pollution shadow performance relationship surface to obtain an initial cleaning strategy that takes into account the impact of shadows, including: using the pollution shadow performance relationship surface to calculate the predicted potential power generation of each photovoltaic component, and based on the potential power generation, performing a cost-benefit analysis on the photovoltaic components to obtain a preliminary cleaning priority list; based on the light intensity data in the component performance status information, using the pollution shadow performance relationship surface to predict the power generation loss data of the pollution shadow performance of the photovoltaic component within a preset time interval, and determining the optimal cleaning time window based on the power generation loss data of the pollution shadow performance and a preset historical load curve of the power station; based on the component pollution degree information and the preset historical cleaning data, estimating the cleaning resource requirements of each photovoltaic component, and based on the preset cleaning priority, the optimal cleaning time window and the cleaning resource requirements, performing a preliminary cleaning task sorting on the photovoltaic components, and adjusting the preliminary task sorting based on the component type and installation method corresponding to each photovoltaic component to obtain a preliminary cleaning strategy that takes into account the impact of shadows.
[0009] Optionally, in a fourth implementation method of the first aspect of the present invention, the photovoltaic cleaning robot includes at least one photovoltaic cleaning robot, and the optimization calculation of the cleaning motion parameters of the initial cleaning strategy to obtain cleaning trajectory parameters that minimize the impact of shadows includes: based on the preliminary cleaning strategy, constructing a weighted directed graph representing the cleaning task, and using the light intensity data in the component performance status information and the contaminated shadow performance relationship surface to calculate the dynamic shadow trajectory of the photovoltaic cleaning robot during the cleaning process; fusing the dynamic shadow trajectory and the weighted directed graph, and performing a cleaning path search on the fusion result to obtain a preliminary cleaning path; based on the component type corresponding to the photovoltaic component according to the component contamination degree information, calculating locally adaptive photovoltaic cleaning parameters for each cleaning point in the preliminary cleaning path; based on the preset power station power generation efficiency curve and the position and posture data of each of the photovoltaic cleaning robots, optimizing the secondary cleaning task allocation of the photovoltaic cleaning parameters corresponding to the photovoltaic cleaning robots to obtain cleaning trajectory parameters that minimize the impact of shadows.
[0010] Optionally, in a fifth implementation of the first aspect of the present invention, based on the cleaning trajectory parameters, the cleaning module is used to perform a cleaning operation on the photovoltaic component to be cleaned to obtain an initial real-time cleaning effect and initial power generation performance data, including: based on the cleaning trajectory parameters, the cleaning module is used to perform a cleaning operation on the photovoltaic component to be cleaned, and the cleaning position data, parameter execution data, real-time power data of the corresponding cleaned photovoltaic component and local area images before and after cleaning are collected during the cleaning operation; based on the real-time power data, a cleaning power curve during the cleaning of the photovoltaic component is calculated, and image cleaning detection is performed on the local area image to obtain real-time cleaning effect data, and based on the cleaning power curve and the component output power, the real-time power generation performance improvement data of the photovoltaic component is calculated; based on the real-time cleaning effect data and the real-time power generation performance improvement data, a cleaning efficiency index of the photovoltaic component is calculated, and based on the cleaning efficiency index, the cleaning position data, the parameter execution data, the real-time cleaning effect data and the real-time power generation performance improvement data are integrated to generate comprehensive initial real-time cleaning effect and initial power generation performance data.
[0011] Optionally, in a sixth implementation of the first aspect of the present invention, the cleaning effect of the initial cleaning result is evaluated based on the initial power generation performance data to obtain an initial cleaning evaluation result, including: based on the initial cleaning result, calculating the percentage of improvement in cleanliness of the component surface before and after cleaning of the photovoltaic component, and based on the percentage of improvement in cleanliness and the cleaning power curve, calculating the increased power generation caused by cleaning the photovoltaic component; based on the contaminated shadow performance relationship surface and the cleaning trajectory parameters, estimating the power generation loss data caused by the shadow in the initial cleaning process, and combining the percentage of improvement in cleanliness, the increased power generation and the power generation loss data to obtain the initial cleaning evaluation result.
[0012] The second aspect of the present invention provides a cleaning control device of a photovoltaic cleaning robot, the cleaning control device of the photovoltaic cleaning robot comprising: a preprocessing module, used to collect multi-source perception data of the photovoltaic components to be cleaned through the perception module, and preprocess the multi-source perception data to obtain component performance status information and component pollution degree information; a cleaning evaluation module, used to perform a cross-analysis of multiple power generation influencing parameters on the component performance status information and the component pollution degree information, to obtain a pollution shadow performance relationship surface of the robot shadow's influence on component power generation, and perform a multi-objective cleaning evaluation on the pollution shadow performance relationship surface to obtain a cleaning evaluation considering the shadow. an initial cleaning strategy for minimizing the influence of shadows; a cleaning operation module, used to optimize the calculation of cleaning motion parameters for the initial cleaning strategy, obtain cleaning trajectory parameters that minimize the influence of shadows, and based on the cleaning trajectory parameters, use the cleaning module to perform cleaning operations on the photovoltaic components to be cleaned, and obtain initial cleaning effects and initial power generation performance data; an adjustment module, used to evaluate the cleaning effects of the initial cleaning results based on the initial power generation performance data, obtain initial cleaning evaluation results, and adjust the cleaning trajectory parameters based on the initial cleaning evaluation results to generate a final cleaning result of the photovoltaic cleaning robot on the photovoltaic components to be cleaned.
[0013] The third aspect of the present invention provides a cleaning control device of a photovoltaic cleaning robot, comprising: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor calls the instructions in the memory so that the cleaning control device of the photovoltaic cleaning robot performs each step of the above-mentioned cleaning control method of the photovoltaic cleaning robot.
[0014] A fourth aspect of the present invention provides a computer-readable storage medium, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes each step of the above-mentioned cleaning control method of the photovoltaic cleaning robot.
[0015] The cleaning control method, device, equipment and storage medium of the above-mentioned photovoltaic cleaning robot. In the embodiment of the present invention, the multi-source sensing data of the photovoltaic components to be cleaned are collected by the sensing module, and the multi-source sensing data are preprocessed to obtain the component performance status information and the component contamination degree information; the component performance status information and the component contamination degree information are cross-analyzed by multiple power generation influencing parameters to obtain the contamination shadow performance relationship surface of the robot shadow on the component power generation, and the contamination shadow performance relationship surface is subjected to multi-objective cleaning evaluation to obtain the initial cleaning strategy considering the shadow influence; the initial cleaning strategy is subjected to the optimization calculation of the cleaning motion parameters to obtain the cleaning trajectory parameters that minimize the shadow influence, and based on the cleaning trajectory parameters, the cleaning module is used to perform the cleaning operation on the photovoltaic components to be cleaned to obtain the initial cleaning effect and the initial power generation performance data; based on the initial power generation performance data, the cleaning effect of the initial cleaning result is evaluated to obtain the initial cleaning evaluation result, and based on the initial cleaning evaluation result, the cleaning trajectory parameters are adjusted to generate the final cleaning result of the photovoltaic cleaning robot for the photovoltaic components to be cleaned. Compared with the prior art, the present application collects multi-source sensing data, and cross-analyzes the pre-processed multi-source sensing data for multiple power generation influencing parameters, obtains the pollution shadow performance relationship surface of the robot shadow on the component power generation, and then performs a multi-objective cleaning evaluation on the pollution shadow performance relationship surface, obtains the initial cleaning strategy that takes into account the shadow influence, and then optimizes the calculation of the cleaning motion parameters of the initial cleaning strategy, obtains the cleaning trajectory parameters that minimize the shadow influence to control the cleaning work of the photovoltaic cleaning robot, and evaluates the cleaning effect of the initial cleaning results, and adjusts the corresponding cleaning trajectory parameters to achieve the final cleaning of the photovoltaic components to be cleaned. Thereby, while ensuring a good cleaning effect on the photovoltaic components, the impact of robot cleaning on photovoltaic power generation efficiency is reduced.
[0016] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.
[0017] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1 Schematic diagram of a first embodiment of a cleaning control method of a photovoltaic cleaning robot in an embodiment of the present invention; Figure 2 A schematic diagram of an embodiment of a cleaning control device of a photovoltaic cleaning robot in an embodiment of the present invention; Figure 3It is a schematic diagram of an embodiment of a cleaning control device of a photovoltaic cleaning robot in an embodiment of the present invention. DETAILED DESCRIPTION
[0019] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. 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.
[0020] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device end including a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or device ends.
[0021] To facilitate understanding of this embodiment, the specific process of the embodiment of the present invention is described below. Figure 1 , the first embodiment of the cleaning control method of the photovoltaic cleaning robot in the embodiment of the present invention includes: 101. Collect multi-source sensing data of the photovoltaic components to be cleaned through the sensing module, and pre-process the multi-source sensing data to obtain component performance status information and component contamination degree information; The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.
[0022] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.
[0023] In this embodiment, the multi-source perception data here includes current and voltage data of photovoltaic components, light intensity data, component surface image data, robot position data and component real-time power generation data. In addition, the types of multi-source perception data can be increased according to the needs of the photovoltaic cleaning robot. The current and voltage data are subjected to wavelet transformation to obtain the current and voltage data with high-frequency noise removed, and based on the current and voltage data, the component output power before cleaning of the photovoltaic component is calculated, and the light intensity data is normalized to obtain standardized light intensity data, and Gaussian blur and edge detection are performed on the component surface image to obtain a pollution distribution image after edge enhancement, and Kalman filter fusion is performed on the robot position data to obtain position and posture data, and the power generation efficiency is calculated for the standardized light intensity data and the component output power using a preset simulation power generation model to obtain a component power generation efficiency coefficient; the current and voltage data with high-frequency noise removed, the component output power, the standardized light intensity data, the pollution distribution map, the position and posture data, and the component power generation efficiency coefficient are timestamp aligned and interpolated to obtain time-synchronized component performance status information; the pollution coverage and pollution degree of the pollution distribution image are calculated to obtain component pollution degree information.
[0024] In practical applications, first, the multi-source perception data of the photovoltaic components to be cleaned are collected through the perception module, such as using the high-precision current transformer and voltage transformer in the perception module to collect the current and voltage data of the photovoltaic components at a sampling frequency of at least 100 Hz; using the thermocouple array in the perception module to collect the surface temperature distribution data of the photovoltaic components at a sampling frequency of at least 1 Hz; using the light intensity sensor in the perception module to collect the incident light intensity data at a sampling frequency of at least 1 Hz; using the camera in the perception module to collect the component surface image data of the photovoltaic components, using the high-precision GPS module in the perception module to collect the real-time position data of the cleaning robot at a sampling frequency of at least 10 Hz; using the acceleration sensor and gyroscope in the perception module to collect the posture data of the cleaning robot at a sampling frequency of at least 100 Hz, using the perception module to record the real-time power generation data of the photovoltaic components, and all the collected data are transmitted and summarized in real time through the 5G network to obtain multi-source perception data.
[0025] Next, select the appropriate wavelet basis function and decomposition layer number, perform multi-scale decomposition on the current and voltage data, process the high-frequency coefficients by the soft threshold method, and finally reconstruct the signal to obtain the denoised current and voltage data. Based on the denoised current and voltage data, use the power calculation formula P=V*I to calculate the output power of the photovoltaic module before cleaning (in addition, for photovoltaic systems, the maximum power point (MPP) is usually concerned, and the perturbation observation method (P&) or incremental conductance method (INC) can be used to track the MPP); and use the min-max normalization method to normalize the light intensity data to obtain standardized light intensity data. For example, if the original light intensity data range is 100-1200 W / m2, after normalization, all data will be mapped to the range of 0-1; and Gaussian blur and edge detection are performed on the component surface image to obtain the pollution distribution image after edge enhancement; and Kalman filter fusion of the robot position data is to obtain more accurate position and posture data. For example, when fusing GPS and IMU data, these formulas are used to estimate the position and posture of the robot, significantly improving the positioning accuracy; and the power generation efficiency is calculated for the standardized light intensity data and component output power using a preset simulation power generation model (such as a single diode model). By comparing the actual output power with the theoretical maximum power, the calculation is obtained. Component power generation efficiency coefficient. Then, the cubic spline interpolation method is used to align and interpolate the timestamps of all processed data to ensure the continuity and smoothness of the data in the time dimension, so as to obtain the time-synchronized component performance status information; then the pollution coverage and pollution degree of the pollution distribution image are calculated to obtain the component pollution degree information, where the pollution coverage is calculated by the pixel ratio after image segmentation, and the pollution degree is quantified by the gray value distribution of the image. Through this series of data processing and analysis steps, the performance status and pollution of photovoltaic components can be comprehensively and accurately evaluated, providing a solid data foundation for the subsequent formulation and implementation of cleaning strategies.
[0026] 102. Perform a cross analysis of multiple power generation influencing parameters on the component performance status information and the component pollution degree information to obtain a pollution shadow performance relationship surface of the robot shadow's impact on the component power generation, and perform a multi-objective cleaning evaluation on the pollution shadow performance relationship surface to obtain an initial cleaning strategy that takes shadow impact into consideration; In this embodiment, a preset robot geometric surface is used to perform a projection transformation of the position and posture data corresponding to the component performance status information to obtain a dynamic shadow contour of the component surface, and based on the preset multiple shadow states and the dynamic shadow contour, the power generation loss of the component power generation efficiency coefficient in the component performance status information is analyzed to obtain the power generation loss data of the local shadow, and based on the corresponding current and voltage data in the component performance status information, the change relationship of the photovoltaic component under different shadow conditions is determined to generate the electrical correlation between components; a multivariate regression analysis is performed on the component pollution degree information and the power generation loss data of the local shadow to generate a preliminary relationship surface, and based on the electrical correlation and the preliminary relationship surface, the power generation characteristics of the photovoltaic component are analyzed to obtain a global power generation influence coefficient; using the preset historical shadow performance data, a time series analysis is performed on the global power generation influence coefficient to generate a shadow influence prediction surface considering the time factor, and based on the component pollution degree information, the shadow influence prediction surface is fused with the pollution degree data to obtain a pollution shadow performance relationship surface of the robot shadow on the component power generation. The predicted potential power generation of each photovoltaic component is calculated by using the pollution shadow performance relationship surface, and based on the potential power generation, the photovoltaic components are subjected to a cost-benefit analysis to obtain a preliminary cleaning priority list; based on the light intensity data in the component performance status information, the pollution shadow performance relationship surface is used to predict the power generation loss data of the pollution shadow performance of the photovoltaic component within a preset time interval, and based on the power generation loss data of the pollution shadow performance and a preset power station historical load curve, the optimal cleaning time window is determined; based on the component pollution degree information and the preset historical cleaning data, the cleaning resource demand of each photovoltaic component is estimated, and based on the preset cleaning priority, the optimal cleaning time window and the cleaning resource demand, the photovoltaic components are preliminarily sorted for cleaning tasks, and based on the component type and installation method corresponding to each photovoltaic component, the preliminary task sorting is adjusted to obtain a preliminary cleaning strategy that takes shadow effects into consideration.
[0027] In practical applications, firstly, the preset robot geometric surface is used to perform a projection transformation of the position and posture data corresponding to the component performance status information, and the three-dimensional space is projected to the two-dimensional plane, that is, the three-dimensional model of the robot is projected onto the plane of the photovoltaic component using the perspective projection matrix, and the projection matrix P can be expressed as: P = K[R|t], where K is the camera intrinsic parameter matrix, R is the rotation matrix, and t is the translation vector. Through the projection transformation, the dynamic shadow contour of the component surface is obtained, and based on the preset multiple shadow states and dynamic shadow contours, the single diode model of the photovoltaic component is used, combined with the IV curve characteristics under different shadow coverage, the component power generation efficiency coefficient in the component performance status information is analyzed for power generation loss, and the power generation loss data of the local shadow is calculated. Based on the corresponding current and voltage data in the component performance status information, the IV and PV curves under different shadow coverage are analyzed by using multi-parameter fitting technology, such as the Levenberg-Marquardt algorithm, to determine the change relationship of the photovoltaic component under different shadow conditions, and generate the electrical correlation between the components; and then by using a multivariate nonlinear regression model, the linear regression calculation formula of the multivariate nonlinear regression model is as follows: ; Among them, P_loss is power loss, S is shadow coverage, D is pollution degree, βO to β5 are regression coefficients, ε is error term, thus multivariate nonlinear regression model is used to perform multivariate regression analysis on component pollution degree information and power generation loss data of local shadow, generate preliminary relationship surface, and analyze power generation characteristics of photovoltaic components based on electrical correlation and preliminary relationship surface, considering series-parallel topology of components and the influence of bypass diode (i.e., by using circuit simulation software (such as SPICE) or custom numerical solution algorithm to simulate the behavior of the entire photovoltaic array), so as to obtain global power generation influence coefficient; then use advanced time Sequence analysis methods, such as the ARIMA (autoregressive integrated moving average) model or the LSTM (long short-term memory) neural network, use preset historical shadow performance data to perform time series analysis on the global power generation impact coefficient to capture the seasonal changes and long-term trends of shadow impact, generate a shadow impact prediction surface that takes time factors into account, and use fuzzy logic methods to define the membership function of shadow impact and pollution level based on component pollution level information, and then obtain the comprehensive performance impact through the fuzzy inference system, so as to evaluate the fusion of pollution level data on the shadow impact prediction surface and obtain the pollution shadow performance relationship surface of the robot shadow on component power generation. The pollution shadow performance relationship surface not only takes into account the dynamic characteristics of the robot shadow, but also includes the impact of pollution level, and also incorporates the results of time series analysis.
[0028] Next, the pollution-shadow-performance relationship surface obtained earlier is used to calculate the power generation under the current pollution state and the assumed cleaning state. In the calculation process, the daily and seasonal changes in light intensity also need to be considered, which is achieved by introducing the standard solar irradiance model to calculate the predicted increased potential power generation of each photovoltaic module after cleaning. Based on the calculated potential power generation, the trade-off between cleaning cost and power generation income is considered, and the net present value (NPV) method is used to evaluate the economic benefits of cleaning each module. A cost-benefit analysis is performed on the photovoltaic modules to obtain a preliminary cleaning priority list. The higher the NPV, the higher the cleaning priority of the module. Based on the light intensity data in the module performance status information, the Monte Carlo simulation method is used to generate a large number of random samples of light intensity and shadow coverage. The corresponding power generation loss is then calculated through the pollution-shadow-performance relationship surface, thereby obtaining the pollution-shadow-performance relationship within the preset time interval. The power generation loss data of the pollution shadow performance is obtained, and a dynamic programming algorithm is used to combine the power generation loss data of the pollution shadow performance and the preset historical load curve of the power station, that is, a day is divided into multiple time periods, and the cleaning decision of each time period will affect the power generation and load balance of the subsequent time period to determine the optimal cleaning time window; then based on the component pollution degree information and the preset historical cleaning data, a regression model is established, and factors such as pollution degree, component area, historical cleaning time, etc. are used as independent variables, and the amount of water, detergent dosage and time required for cleaning are used as dependent variables. These coefficients are estimated by the least squares method to predict the specific resource requirements of each component, thereby estimating the cleaning resource requirements of each photovoltaic component, and based on the preset cleaning priority, optimal cleaning time window and cleaning resource requirements, a multi-objective optimization algorithm such as a genetic algorithm or a particle swarm optimization algorithm is used to perform a preliminary cleaning task sorting for the photovoltaic components (that is, define a comprehensive objective function: F= w1f1+ w2f2+ w3f3+ w4f4, where f1 represents the increase in power generation, f2 represents resource utilization efficiency, f3 represents the time window matching degree, f4 represents the minimization of shadow impact, and w1 to w4 are the corresponding weights. By optimizing this objective function, a cleaning task sequence that balances various factors is obtained), and adjustments are made according to the component type and installation method corresponding to each photovoltaic module, fully considering the impact of the component type and installation method (for example, for bifacial modules, it may be necessary to consider the cleaning requirements on the back; for tracking-mounted modules, it is necessary to coordinate cleaning operations and tracking actions, and these adjustments can be achieved by adding constraints or modifying the objective function), thereby obtaining a preliminary cleaning strategy that takes shadow impact into account. This strategy can not only maximize the economic benefits of the cleaning operation, but also minimize the power generation loss during the cleaning process.
[0029] 103. Optimize and calculate the cleaning motion parameters of the initial cleaning strategy to obtain cleaning trajectory parameters that minimize the shadow effect, and based on the cleaning trajectory parameters, use the cleaning module to perform a cleaning operation on the photovoltaic components to be cleaned to obtain initial cleaning effects and initial power generation performance data; In this embodiment, based on the preliminary cleaning strategy, a weighted directed graph representing the cleaning task is constructed, and the dynamic shadow trajectory of the photovoltaic cleaning robot during the cleaning process is calculated using the light intensity data in the component performance status information and the contaminated shadow performance relationship surface; the dynamic shadow trajectory and the weighted directed graph are fused, and a cleaning path search is performed on the fusion result to obtain a preliminary cleaning path; based on the component type corresponding to the photovoltaic component according to the component contamination level information, local adaptive photovoltaic cleaning parameters are calculated for each cleaning point in the preliminary cleaning path; based on the preset power station power generation efficiency curve and the position and posture data of each photovoltaic cleaning robot, the photovoltaic cleaning parameters corresponding to the photovoltaic cleaning robot are optimized for secondary cleaning task allocation to obtain cleaning trajectory parameters that minimize the impact of shadows. Based on the cleaning trajectory parameters, the cleaning module is used to perform a cleaning operation on the photovoltaic component to be cleaned, and the cleaning position data, parameter execution data, real-time power data of the corresponding cleaned photovoltaic component and local area images before and after cleaning are collected during the cleaning operation; based on the real-time power data, the cleaning power curve during the cleaning of the photovoltaic component is calculated, and image cleaning detection is performed on the local area image to obtain real-time cleaning effect data, and based on the cleaning power curve and the component output power, the real-time power generation performance improvement data of the photovoltaic component is calculated; based on the real-time cleaning effect data and the real-time power generation performance improvement data, the cleaning efficiency index of the photovoltaic component is calculated, and based on the cleaning efficiency index, the cleaning position data, the parameter execution data, the real-time cleaning effect data and the real-time power generation performance improvement data are integrated to generate comprehensive initial real-time cleaning effect and initial power generation performance data.
[0030] In practical applications, first, based on the preliminary cleaning strategy obtained above, a weighted graph representing the cleaning task is constructed. This graph can be represented by G=(V,E,W), where V is a vertex set representing each photovoltaic component that needs to be cleaned, E is an edge set representing the possible moving path of the cleaning robot, and W is a weight set, which can contain weights of multiple dimensions, such as moving distance, cleaning priority, expected cleaning time, etc. The construction of the graph is achieved by using an adjacency matrix or an adjacency list. Considering that the graph may be sparse, an adjacency list may be more appropriate. The light intensity data in the component performance status information and the pollution shadow performance relationship surface are used. The solar position algorithm (such as the SPA algorithm) is used to calculate the solar azimuth and altitude angle at a given time and geographic location. Then, combined with the robot's geometric model, the shape and position of the shadow are calculated through projection transformation, thereby calculating the dynamic shadow trajectory of the photovoltaic cleaning robot during the cleaning process; the dynamic shadow trajectory and the weighted directed graph are then fused, that is, by adding a time-related shadow influence factor to each edge of the graph, for example, by defining a new edge weight: ; in, is the edge weight after considering the shadow effect, w(e) is the original edge weight, α is the balancing factor, and ShadowImpact(e, t) is the shadow effect on edge e at time t; and since searching for the cleaning path of the fused graph is the next key step, considering the dynamics and complexity of the problem, the improved A algorithm is used to find the preliminary cleaning path to find a path that balances the cleaning efficiency and the shadow effect; and then based on the component contamination level information and the component type corresponding to the photovoltaic component, a parameter optimization model (such as a multi-objective optimization method) is constructed to simultaneously consider the cleaning effect, resource consumption and time efficiency to calculate the local adaptive photovoltaic cleaning parameters for each cleaning point in the preliminary cleaning path; and then based on the preset power station power generation efficiency curve and the position and posture data of each photovoltaic cleaning robot, the photovoltaic cleaning parameters corresponding to the photovoltaic cleaning robot are optimized for secondary cleaning task allocation (that is, by defining the state space S (including robot position, component cleaning state, etc.), action space A (cleaning parameter adjustment, moving direction, etc.), and reward function R (considering cleaning effect and shadow effect). Then use deep Q learning (DQN) or policy gradient method to learn the optimal strategy) to obtain the cleaning trajectory parameters that minimize the shadow effect. Finally, an optimized cleaning trajectory and parameters that fully consider cleaning efficiency, shadow effects, component characteristics, and system dynamics are obtained. This result can not only guide the cleaning robot to complete the cleaning task efficiently, but also minimize the power generation loss during the cleaning process.
[0031] Next, based on the cleaning trajectory parameters obtained by the previous optimization, the cleaning module is used to clean the photovoltaic components to be cleaned. During the cleaning process, multiple data need to be collected in real time, including cleaning position data, parameter execution data, real-time power data of the corresponding cleaned photovoltaic components, and local area images before and after cleaning. The cleaning position data can be obtained through the GPS module and inertial measurement unit (IMU) on the robot. By fusing these two data sources through the Kalman filter algorithm, high-precision position information can be obtained; parameter execution data include cleaning pressure, water flow, cleaning speed, etc., which can be collected in real time through various sensors; real-time power data can be obtained through a micro power analyzer installed on the photovoltaic component, which can record current, voltage and power data at a high sampling rate; local area images before and after cleaning can be captured by a high-resolution camera installed on the robot. Based on the real-time electric energy data collected, the sliding window method is used to calculate the half-mean power in each time window, and then these half-mean power values are connected to form a smooth power curve, so as to calculate the cleaning power curve in the cleaning of the photovoltaic module, and the local area images before and after cleaning need to be image cleaned. The image is divided into clean area and unclean area by using an image segmentation algorithm (such as U-Net), and the area ratio and cleanliness of the clean area are calculated (that is, quantified by analyzing the brightness and texture features of the clean area), so as to obtain a real-time cleaning effect index to obtain real-time cleaning effect data; based on the cleaning power curve and the previously recorded component output power, the power output before and after cleaning is compared to calculate the cleaning effect. The real-time power generation performance improvement data of the PV modules is calculated to intuitively reflect the immediate performance improvement brought by the cleaning operation; then, based on the real-time cleaning effect data and the real-time power generation performance improvement data, the cleaning efficiency index of the PV modules is calculated to reflect the cleaning effect and performance improvement per unit time and per unit resource consumption, which can comprehensively evaluate the efficiency of the cleaning operation, and based on the cleaning efficiency index, by constructing a multidimensional data structure, in which each dimension represents a key indicator, to integrate the cleaning position data, parameter execution data, real-time cleaning effect data and real-time power generation performance improvement data, to generate comprehensive initial real-time cleaning effect and initial power generation performance data, so as to evaluate the effect and efficiency of the cleaning operation in real time, and provide an important basis for the subsequent adjustment of the cleaning strategy.
[0032] 104. Based on the initial power generation performance data, the cleaning effect of the initial cleaning result is evaluated to obtain the initial cleaning evaluation result, and based on the initial cleaning evaluation result, the cleaning trajectory parameters are adjusted to generate the final cleaning result of the photovoltaic cleaning robot for the photovoltaic component to be cleaned.
[0033] In this embodiment, based on the initial cleaning result, the percentage of improvement in cleanliness of the component surface before and after cleaning of the photovoltaic component is calculated, and based on the percentage of improvement in cleanliness and the cleaning power curve, the increased power generation caused by cleaning of the photovoltaic component is calculated; based on the contamination shadow performance relationship surface and the cleaning trajectory parameters, the power generation loss data caused by the shadow during the initial cleaning process is estimated, and the initial cleaning evaluation result is obtained by combining the percentage of improvement in cleanliness, the increased power generation and the power generation loss data.
[0034] In practical applications, firstly, based on the initial cleaning results, image processing techniques such as grayscale analysis and texture feature extraction are used to quantify the cleanliness of the component surface, so as to calculate the percentage of cleanliness improvement of the component surface before and after cleaning of the photovoltaic component, where this indicator can be defined by the average brightness, contrast or specific texture features of the image; and based on the percentage of cleanliness improvement and the cleaning power curve, the area difference between the power curves before and after cleaning is calculated by numerical integration (where the numerical integration interval can be a typical power generation cycle, such as one day or one week), so as to calculate the increased power generation caused by cleaning of the photovoltaic component; and then based on the pollution shadow performance relationship surface and the cleaning trajectory parameters, the movement of the cleaning robot on the cleaning trajectory is simulated, and the pollution shadow performance relationship surface is combined to estimate the power generation loss data caused by the shadow during the initial cleaning process; and then a comprehensive score is constructed The model is used to integrate the data of cleanliness improvement percentage, power generation improvement and power generation loss to obtain the initial cleaning evaluation result, where the initial cleaning evaluation result reflects the overall effect of the cleaning operation, while taking into account the cleanliness improvement, power generation increase and shadow loss; based on the initial cleaning evaluation result, the gradient descent or evolutionary algorithm is used to optimize the cleaning trajectory parameters (i.e., by iteratively adjusting the cleaning trajectory parameters (such as cleaning speed, water pressure, number of cleanings, etc.), the objective function is recalculated after each adjustment until the convergence condition is reached or the preset number of iterations is reached, and in each iteration, the cleaning trajectory parameters are updated to obtain an optimized cleaning trajectory parameter set, which can achieve a good balance between cleaning effect and cleaning cost), and then the optimized parameter set is used to perform the last cleaning operation to generate the final cleaning result of the photovoltaic cleaning robot for the photovoltaic modules to be cleaned. In this way, while improving the power generation efficiency, the power generation loss and resource consumption in the cleaning process are minimized.
[0035] In the embodiment of the present invention, by collecting multi-source sensing data and performing cross analysis of multiple power generation influencing parameters on the pre-processed multi-source sensing data, a contaminated shadow performance relationship surface of the robot shadow's influence on the component power generation is obtained, and then a multi-objective cleaning evaluation is performed on the contaminated shadow performance relationship surface to obtain an initial cleaning strategy that takes into account the shadow influence, so that the initial cleaning strategy performs an optimization calculation of the cleaning motion parameters, obtains the cleaning trajectory parameters that minimize the shadow influence to control the cleaning work of the photovoltaic cleaning robot, and evaluates the cleaning effect of the initial cleaning result, and adjusts the corresponding cleaning trajectory parameters to achieve the final cleaning of the photovoltaic component to be cleaned. Thereby, while ensuring a good cleaning effect on the photovoltaic component, the impact of the robot cleaning on the photovoltaic power generation efficiency is reduced.
[0036] The above describes the cleaning control method of the photovoltaic cleaning robot in the embodiment of the present invention. The following describes the cleaning control device of the photovoltaic cleaning robot in the embodiment of the present invention. Figure 2 In one embodiment of the present invention, a cleaning control device of a photovoltaic cleaning robot includes: The preprocessing module 201 is used to collect multi-source sensing data of the photovoltaic components to be cleaned through the sensing module, and preprocess the multi-source sensing data to obtain component performance status information and component contamination degree information; The cleaning evaluation module 202 is used to perform a cross analysis of multiple power generation influencing parameters on the component performance status information and the component pollution degree information, obtain a pollution shadow performance relationship surface of the robot shadow's influence on the component power generation, and perform a multi-objective cleaning evaluation on the pollution shadow performance relationship surface to obtain an initial cleaning strategy that takes shadow influence into consideration; A cleaning operation module 203 is used to optimize the calculation of cleaning motion parameters for the initial cleaning strategy to obtain cleaning trajectory parameters that minimize the influence of shadows, and based on the cleaning trajectory parameters, use the cleaning module to perform a cleaning operation on the photovoltaic assembly to be cleaned to obtain initial cleaning effect and initial power generation performance data; The adjustment module 204 is used to evaluate the cleaning effect of the initial cleaning result based on the initial power generation performance data to obtain an initial cleaning evaluation result, and adjust the cleaning trajectory parameters based on the initial cleaning evaluation result to generate the final cleaning result of the photovoltaic cleaning robot on the photovoltaic component to be cleaned.
[0037] In the embodiment of the present invention, by collecting multi-source sensing data and performing cross analysis of multiple power generation influencing parameters on the pre-processed multi-source sensing data, a contaminated shadow performance relationship surface of the robot shadow's influence on the component power generation is obtained, and then a multi-objective cleaning evaluation is performed on the contaminated shadow performance relationship surface to obtain an initial cleaning strategy that takes into account the shadow influence, so that the initial cleaning strategy performs an optimization calculation of the cleaning motion parameters, obtains the cleaning trajectory parameters that minimize the shadow influence to control the cleaning work of the photovoltaic cleaning robot, and evaluates the cleaning effect of the initial cleaning result, and adjusts the corresponding cleaning trajectory parameters to achieve the final cleaning of the photovoltaic component to be cleaned. Thereby, while ensuring a good cleaning effect on the photovoltaic component, the impact of the robot cleaning on the photovoltaic power generation efficiency is reduced.
[0038] above Figure 2 The cleaning control device of the photovoltaic cleaning robot in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The cleaning control device of the photovoltaic cleaning robot in the embodiment of the present invention is described in detail from the perspective of hardware processing.
[0039] Figure 3 3 is a schematic diagram of the structure of a cleaning control device of a photovoltaic cleaning robot provided by an embodiment of the present invention. The cleaning control device 300 of the photovoltaic cleaning robot may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 310 (for example, one or more processors) and a memory 320, and one or more storage media 330 (for example, one or more mass storage devices) storing application programs 333 or data 332. Among them, the memory 320 and the storage medium 330 can be short-term storage or permanent storage. The program stored in the storage medium 330 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the cleaning control device 300 of the photovoltaic cleaning robot. Furthermore, the processor 310 may be configured to communicate with the storage medium 330 to execute a series of instruction operations in the storage medium 330 on the cleaning control device 300 of the photovoltaic cleaning robot.
[0040] The cleaning control device 300 of the photovoltaic cleaning robot may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating systems 331, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It can be understood by those skilled in the art that Figure 3The cleaning control device structure of the photovoltaic cleaning robot shown does not constitute a limitation on the cleaning control device of the photovoltaic cleaning robot, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.
[0041] The present invention also provides a cleaning control device for a photovoltaic cleaning robot, wherein the computer device includes a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes each step of the cleaning control method for the photovoltaic cleaning robot in the above-mentioned embodiments.
[0042] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. Instructions are stored in the computer-readable storage medium. When the instructions are executed on a computer, the computer executes the various steps of the cleaning control method of the photovoltaic cleaning robot.
[0043] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0044] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.
[0045] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.
[0046] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A cleaning control method for a photovoltaic cleaning robot, the photovoltaic cleaning robot comprising a sensing module and a cleaning module, characterized in that: The cleaning control method of the photovoltaic cleaning robot comprises: The sensing module collects multi-source sensing data of the photovoltaic components to be cleaned, and pre-processes the multi-source sensing data to obtain component performance status information and component contamination degree information; Performing a cross analysis of multiple power generation influencing parameters on the component performance status information and the component contamination degree information to obtain a contamination shadow performance relationship surface of the robot shadow's influence on component power generation, and performing a multi-objective cleaning evaluation on the contamination shadow performance relationship surface to obtain an initial cleaning strategy that takes shadow influence into account; Optimizing and calculating the cleaning motion parameters of the initial cleaning strategy to obtain cleaning trajectory parameters that minimize the influence of shadows, and based on the cleaning trajectory parameters, using the cleaning module to perform a cleaning operation on the photovoltaic assembly to be cleaned to obtain initial cleaning effect and initial power generation performance data; Based on the initial power generation performance data, the cleaning effect of the initial cleaning result is evaluated to obtain an initial cleaning evaluation result, and based on the initial cleaning evaluation result, the cleaning trajectory parameters are adjusted to generate a final cleaning result of the photovoltaic cleaning robot on the photovoltaic component to be cleaned.
2. The cleaning control method of the photovoltaic cleaning robot according to claim 1, characterized in that: The multi-source sensing data includes current and voltage data of photovoltaic modules, light intensity data, module surface image data, robot position data and module real-time power generation data. The multi-source sensing data is preprocessed to obtain module performance status information and module contamination degree information, including: Performing wavelet transform on the current and voltage data to obtain current and voltage data with high-frequency noise removed, and calculating the component output power of the photovoltaic component before cleaning based on the current and voltage data, normalizing the light intensity data to obtain standardized light intensity data, performing Gaussian blur and edge detection on the component surface image to obtain a pollution distribution image after edge enhancement, performing Kalman filtering fusion on the robot position data to obtain position posture data, and using a preset simulation power generation model to calculate the power generation efficiency of the standardized light intensity data and the component output power to obtain a component power generation efficiency coefficient; Performing timestamp alignment and interpolation on the current and voltage data with high-frequency noise removed, the component output power, the standardized light intensity data, the pollution distribution map, the position and posture data, and the component power generation efficiency coefficient to obtain time-synchronized component performance status information; The pollution coverage and pollution degree of the pollution distribution image are calculated to obtain component pollution degree information.
3. The cleaning control method of the photovoltaic cleaning robot according to claim 2, characterized in that: The cross analysis of multiple power generation influencing parameters on the component performance status information and the component pollution degree information is performed to obtain a pollution shadow performance relationship surface of the robot shadow's influence on component power generation, including: Using a preset robot geometric surface to perform position and posture projection transformation on the corresponding position and posture data in the component performance status information, a dynamic shadow profile of the component surface is obtained, and based on the preset multiple shadow states and the dynamic shadow profile, the power generation loss of the component power generation efficiency coefficient in the component performance status information is analyzed to obtain the power generation loss data of the local shadow, and based on the corresponding current and voltage data in the component performance status information, the change relationship of the photovoltaic component under different shadow conditions is determined to generate the electrical correlation between the components; Performing a multivariate regression analysis on the component pollution degree information and the power generation loss data of the local shadow to generate a preliminary relationship surface, and performing a power generation characteristic analysis on the photovoltaic component based on the electrical correlation and the preliminary relationship surface to obtain a global power generation influence coefficient; The preset historical shadow performance data is used to perform a time series analysis on the global power generation impact coefficient to generate a shadow impact prediction surface that takes time factors into account. Based on the component pollution level information, the shadow impact prediction surface is fused with the pollution level data to obtain a pollution-shadow-performance relationship surface of the robot shadow's impact on component power generation.
4. The cleaning control method of the photovoltaic cleaning robot according to claim 1, characterized in that: The multi-objective cleaning evaluation is performed on the pollution shadow performance relationship surface to obtain an initial cleaning strategy that takes shadow influence into consideration, including: Using the pollution shadow performance relationship surface, the predicted potential power generation of each photovoltaic module is calculated, and based on the potential power generation, a cost-benefit analysis of the photovoltaic modules is performed to obtain a preliminary cleaning priority list; Based on the light intensity data in the component performance status information, the pollution shadow performance relationship surface is used to predict the power generation loss data of the pollution shadow performance of the photovoltaic component within a preset time interval, and based on the power generation loss data of the pollution shadow performance and a preset power station historical load curve, an optimal cleaning time window is determined; Based on the component contamination level information and preset historical cleaning data, the cleaning resource requirements of each photovoltaic component are estimated, and based on the preset cleaning priority, the optimal cleaning time window and the cleaning resource requirements, the photovoltaic components are preliminarily sorted for cleaning tasks, and based on the component type and installation method corresponding to each photovoltaic component, the preliminary task sorting is adjusted to obtain a preliminary cleaning strategy that takes into account the impact of shadows.
5. The cleaning control method of the photovoltaic cleaning robot according to claim 4, characterized in that: The photovoltaic cleaning robot includes at least one photovoltaic cleaning robot, and the optimization calculation of the cleaning motion parameters of the initial cleaning strategy to obtain the cleaning trajectory parameters that minimize the shadow effect includes: Based on the preliminary cleaning strategy, a weighted directed graph representing the cleaning task is constructed, and the dynamic shadow trajectory of the photovoltaic cleaning robot during the cleaning process is calculated using the light intensity data in the component performance status information and the contamination shadow performance relationship surface; The dynamic shadow trajectory and the weighted directed graph are fused, and a clean path search is performed on the fused result to obtain a preliminary clean path; Based on the component contamination level information and the component type corresponding to the photovoltaic component, a local adaptive photovoltaic cleaning parameter is calculated for each cleaning point in the preliminary cleaning path; Based on the preset power station power generation efficiency curve and the position and posture data of each photovoltaic cleaning robot, the photovoltaic cleaning parameters corresponding to the photovoltaic cleaning robots are optimized for secondary cleaning task allocation to obtain cleaning trajectory parameters that minimize the impact of shadows.
6. The cleaning control method of the photovoltaic cleaning robot according to claim 3, characterized in that: Based on the cleaning trajectory parameters, the cleaning module is used to perform a cleaning operation on the photovoltaic assembly to be cleaned, and an initial real-time cleaning effect and initial power generation performance data are obtained, including: Based on the cleaning trajectory parameters, the cleaning module is used to perform a cleaning operation on the photovoltaic component to be cleaned, and cleaning position data, parameter execution data, real-time power data of the photovoltaic component to be cleaned, and local area images before and after cleaning are collected during the cleaning operation; Based on the real-time electric energy data, a cleaning power curve in the cleaning of the photovoltaic module is calculated, and image cleaning detection is performed on the local area image to obtain real-time cleaning effect data, and based on the cleaning power curve and the module output power, real-time power generation performance improvement data of the photovoltaic module is calculated; Based on the real-time cleaning effect data and the real-time power generation performance improvement data, the cleaning efficiency index of the photovoltaic component is calculated, and based on the cleaning efficiency index, the cleaning position data, the parameter execution data, the real-time cleaning effect data and the real-time power generation performance improvement data are integrated to generate comprehensive initial real-time cleaning effect and initial power generation performance data.
7. The cleaning control method of the photovoltaic cleaning robot according to claim 6, characterized in that: The step of evaluating the cleaning effect of the initial cleaning result based on the initial power generation performance data to obtain an initial cleaning evaluation result includes: Based on the initial cleaning result, calculating the percentage of improvement in cleanliness of the surface of the photovoltaic module before and after cleaning, and based on the percentage of improvement in cleanliness and the cleaning power curve, calculating the increased power generation resulting from the cleaning of the photovoltaic module; Based on the pollution shadow performance relationship surface and the cleaning trajectory parameters, the power generation loss data caused by the shadows during the initial cleaning process is estimated, and the cleanliness improvement percentage, the improved power generation and the power generation loss data are combined to obtain the initial cleaning evaluation result.
8. A cleaning control device for a photovoltaic cleaning robot, characterized in that: The cleaning control device of the photovoltaic cleaning robot comprises: A preprocessing module, used to collect multi-source sensing data of the photovoltaic components to be cleaned through the sensing module, and preprocess the multi-source sensing data to obtain component performance status information and component contamination degree information; A cleaning evaluation module is used to perform a cross-analysis of multiple power generation influencing parameters on the component performance status information and the component contamination degree information, obtain a contamination shadow performance relationship surface of the robot shadow's influence on component power generation, and perform a multi-objective cleaning evaluation on the contamination shadow performance relationship surface to obtain an initial cleaning strategy that takes shadow influence into account; A cleaning operation module, used to optimize and calculate the cleaning motion parameters of the initial cleaning strategy to obtain cleaning trajectory parameters that minimize the shadow effect, and based on the cleaning trajectory parameters, use the cleaning module to perform a cleaning operation on the photovoltaic assembly to be cleaned to obtain initial cleaning effect and initial power generation performance data; An adjustment module is used to evaluate the cleaning effect of the initial cleaning result based on the initial power generation performance data to obtain an initial cleaning evaluation result, and based on the initial cleaning evaluation result, adjust the cleaning trajectory parameters to generate a final cleaning result of the photovoltaic cleaning robot on the photovoltaic component to be cleaned.
9. A cleaning control device for a photovoltaic cleaning robot, characterized in that: The cleaning control device of the photovoltaic cleaning robot comprises: a memory and at least one processor, wherein instructions are stored in the memory; The at least one processor calls the instructions in the memory so that the cleaning control device of the photovoltaic cleaning robot executes each step of the cleaning control method of the photovoltaic cleaning robot as described in any one of claims 1-7.
10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the various steps of the cleaning control method of the photovoltaic cleaning robot as described in any one of claims 1-7 are implemented.
Citation Information
Cited By
Configuration generation method and device of unmanned aerial vehicle cleaning system and unmanned aerial vehicle cleaning system
CN119414732A
Photovoltaic multi-source sensing intelligent cleaning and cooling method and system
CN120128071A
Photovoltaic Multi-Source Sensing Intelligent Cleaning and Cooling Method and System
CN120128071B
Intelligent photovoltaic module self-cleaning device and method
CN120301334A
Photovoltaic panel full-automatic movable cleaning system and method based on timing control
CN120377792A
