Photovoltaic panel adaptive cleaning path planning method and system based on environmental perception

Abstract

This invention provides a method and system for adaptive cleaning path planning of photovoltaic panels based on environmental perception, relating to the field of photovoltaic panel cleaning technology. The method includes calculating a cleanliness value by acquiring historical weather data and status monitoring data, triggering cleaning when the value falls below a dynamic cleaning threshold; optimizing the cleaning path using a multi-pheromone ant colony algorithm and an environmentally adaptive multi-neighborhood simulated annealing algorithm; and dividing the area into sub-regions based on dirt density and enabling multi-robot collaborative cleaning. This invention improves cleaning efficiency, reduces energy consumption, and achieves intelligent and precise photovoltaic panel cleaning.

Description

Technical Field

[0001] This invention relates to photovoltaic panel cleaning technology, and more particularly to a photovoltaic panel adaptive cleaning path planning method and system based on environmental perception. Background Technology

[0002] With the widespread application of photovoltaic power generation technology, the impact of dust accumulation on photovoltaic panel surfaces on power generation efficiency is becoming increasingly prominent. Traditional photovoltaic panel cleaning mainly relies on manual labor, which is not only labor-intensive and inefficient but also poses safety hazards. With the development of automation technology, robotic cleaning systems are gradually being applied to the field of photovoltaic panel cleaning, but existing automated photovoltaic panel cleaning systems still have many shortcomings in cleaning path planning.

[0003] The main problems with current photovoltaic panel cleaning path planning technology include: First, existing cleaning systems lack an adaptive cleaning decision-making mechanism based on environmental perception. Most systems use fixed cycles or simple thresholds to trigger cleaning, failing to fully consider the impact of environmental factors such as weather changes and dust accumulation rates on cleaning timing, leading to wasted cleaning resources or power generation losses due to untimely cleaning. Second, existing cleaning path planning algorithms typically only consider a single optimization objective, such as the shortest path length or the largest coverage area, failing to comprehensively consider multiple dimensions such as cleaning efficiency, energy consumption, and time cost, making it difficult to achieve global optimization in complex environments. Finally, the multi-robot collaborative cleaning control strategy is imperfect, lacking an intelligent task allocation mechanism based on dirt distribution and effective obstacle avoidance strategies, resulting in uneven resource allocation and severe interference between robots during the cleaning process, affecting overall cleaning efficiency and quality. Summary of the Invention

[0004] This invention provides a method and system for adaptive cleaning path planning of photovoltaic panels based on environmental perception, which can solve the problems in the prior art.

[0005] A first aspect of the present invention,

[0006] A method for adaptive cleaning path planning of photovoltaic panels based on environmental awareness is provided, including:

[0007] Historical weather data and photovoltaic panel status monitoring data of the area where the photovoltaic panel is located are obtained, and the cleanliness value of the photovoltaic panel is calculated; when the cleanliness value is lower than the cleaning threshold dynamically calculated based on power generation loss and cleaning cost, a cleaning trigger command is generated.

[0008] Upon receiving the cleaning trigger command, the cleaning path optimization steps are executed, including: setting the optimization objectives as maximizing cleaning efficiency, minimizing energy consumption, and optimizing time cost; dynamically adjusting the weights based on environmental impact and cleaning energy efficiency ratio; optimizing the population using a genetic algorithm based on dirt density encoding; mapping the optimized population to the initial distribution of cleaning efficiency pheromones, energy consumption pheromones, and time cost pheromones; obtaining the global path through iterative calculation using a multi-pheromone ant colony algorithm; and performing local optimization using an environmentally adaptive multi-neighborhood simulated annealing algorithm.

[0009] Based on the optimized cleaning path, multi-robot collaborative cleaning control steps are executed, including: dividing the area into sub-regions based on dirt density and workload balance; using segmented adaptive sliding mode variable structure control for trajectory tracking; determining obstacle avoidance priority based on the percentage of cleaned area and target distance, and performing multi-robot collaborative control.

[0010] In one alternative implementation,

[0011] Historical weather data and photovoltaic panel status monitoring data for the area where the photovoltaic panels are located are obtained, and the cleanliness value of the photovoltaic panels is calculated, including:

[0012] Historical weather data is processed using a sliding time window method. A time decay factor is introduced into the data within each time window for weighted calculation to obtain the historical weather impact value.

[0013] The historical weather impact values ​​and status monitoring data are multiplied by the corresponding weight coefficients and summed to obtain the photovoltaic panel cleanliness value. The initial values ​​of each weight coefficient are determined by fuzzy hierarchical analysis, and the gradient direction of the weights is dynamically adjusted based on the cleanliness value during the calculation process.

[0014] Establish a historical cleaning record database to record the cleanliness value before each cleaning, cleaning cost, and actual power generation loss during the corresponding period. Combine seasonal and environmental factors to calculate the power generation loss caused by the reduction in unit cleanliness.

[0015] A cleaning benefit evaluation function is established based on the unit cleaning cost and the power generation loss caused by the reduction in unit cleanliness. The corresponding cleaning benefit is calculated by traversing different cleanliness values, and the cleanliness value that maximizes the cleaning benefit is determined as the cleaning threshold for the current period. The cleaning benefit evaluation function also considers the cost of cleaning resource scheduling and the time window constraint of cleaning operation.

[0016] In one alternative implementation,

[0017] The weights are dynamically adjusted based on environmental impact and cleaning energy efficiency ratio, including:

[0018] An environmental impact function is established to characterize the effects of temperature, humidity, wind speed, light intensity, and precipitation probability on the cleaning effect. The temperature parameter is calculated using an exponential function to determine the degree of influence of temperature deviation from the optimal cleaning temperature, and the humidity parameter is calculated using an S-shaped response curve to determine the degree of influence of humidity exceeding a critical value. A coupling coefficient between the temperature and humidity parameters is constructed based on the environmental impact function.

[0019] The grid cells are divided based on the density of dirt distribution on the photovoltaic panel surface, and initial weights are set for cleaning efficiency, energy consumption, and time cost.

[0020] An environmental adjustment coefficient is calculated based on the environmental impact function and the coupling coefficient, and the environmental adjustment coefficient is positively correlated with the environmental sensitivity of the grid cell; an energy efficiency ratio adjustment coefficient is calculated based on the cleaning energy efficiency ratio of the grid cell, and the energy efficiency ratio adjustment coefficient is a power function relationship between the current cleaning energy efficiency ratio and the maximum cleaning energy efficiency ratio;

[0021] The product of the environmental adjustment coefficient and the energy efficiency ratio adjustment coefficient is used as the weight adjustment amount, so that the cleaning efficiency weight is negatively correlated with the weight adjustment amount, and the energy consumption weight and the time cost weight are positively correlated with the weight adjustment amount, and the initial weights are dynamically updated.

[0022] In one alternative implementation,

[0023] A genetic algorithm based on dirt density encoding is used to optimize the population, mapping the optimized population to the initial distribution of cleaning efficiency pheromones, energy consumption pheromones, and time cost pheromones. A global path is obtained through iterative calculation using a multi-pheromone ant colony algorithm, and local optimization is performed using an environment-adaptive multi-neighborhood simulated annealing algorithm, including:

[0024] The initial population is initialized and constructed using a genetic algorithm, which employs an encoding strategy based on dirt density weights and an environment-aware adaptive crossover and mutation mechanism.

[0025] The optimization results of the genetic algorithm are used as the initial pheromone distribution of the ant colony algorithm. A multi-pheromone ant colony algorithm based on grid energy efficiency gradient is used to generate the globally optimal cleaning path. The multi-pheromone includes cleaning efficiency pheromone, energy consumption pheromone and time cost pheromone.

[0026] The globally optimal cleaning path is locally optimized using an environment-adaptive multi-neighborhood simulated annealing algorithm. The multi-neighborhood includes reverse neighborhood, exchange neighborhood, insertion neighborhood, and region reconstruction neighborhood. Local path optimization is performed through an environment-aware temperature adjustment mechanism and a dynamic acceptance criterion strategy.

[0027] In one alternative implementation,

[0028] The genetic algorithm employs an encoding strategy based on dirt density weights and an environment-aware adaptive crossover and mutation mechanism, including:

[0029] The characteristics of various types of dirt on the surface of the photovoltaic panel are obtained, the surface of the photovoltaic panel is divided into grid cells, and the dirt density of each grid cell is calculated.

[0030] The chromosome of the genetic algorithm is constructed as a multi-level coding structure of spatial sequence coding, resource allocation coding and density correlation coding, where spatial sequence coding represents the cleaning path, resource allocation coding represents the cleaning parameters, and density correlation coding represents the dirt handling strategy.

[0031] A density weight mapping function is constructed based on the dirt density. The density weight mapping function simultaneously considers the dirt density of the target grid cell and the dirt density distribution of its adjacent grid cells. The population is initialized according to the density weight mapping function.

[0032] Environmental parameters are acquired and environmental impact factors are calculated. Based on the environmental impact factors and the density weight mapping function, the crossover probability and mutation probability of the genetic algorithm are dynamically adjusted. The cleaning path sequence is optimized through crossover operation, and the resource allocation strategy is adjusted through mutation operation.

[0033] The spatial sequence encoding, resource allocation encoding, and density correlation encoding of the best individual in the optimized population are mapped to the initial pheromone distribution matrix of the ant colony algorithm.

[0034] In one alternative implementation,

[0035] The optimization results of the genetic algorithm are used as the initial pheromone distribution for the ant colony algorithm. A multi-pheromone ant colony algorithm based on grid energy efficiency gradient is used to generate the globally optimal cleaning path. The multi-pheromone includes cleaning efficiency pheromone, energy consumption pheromone, and time cost pheromone.

[0036] The optimization result of the genetic algorithm is mapped to a multidimensional pheromone initial distribution, which includes cleaning efficiency pheromone, energy consumption pheromone and time cost pheromone. Each pheromone component is obtained by weighted combination of the sequence information corresponding to the optimal individual and the grid dirt density.

[0037] The grid energy efficiency ratio is calculated based on the ratio of cleaning revenue to energy consumption of the grid cells, and the grid energy efficiency gradient vector is constructed based on the partial derivatives of the grid energy efficiency ratio in the horizontal and vertical directions.

[0038] The heuristic information of each pheromone component is weighted and multiplied to obtain the comprehensive heuristic information. Based on the comprehensive heuristic information, the initial distribution of multidimensional pheromones, and the grid energy efficiency gradient vector, the state transition probability matrix is ​​calculated.

[0039] The ant's movement path is determined based on the state transition probability matrix. Local pheromone updates are performed on the grid cells along the movement path, taking into account the influence of the pheromone evaporation coefficient and the grid energy efficiency ratio. The global optimal path is then enhanced with pheromone based on the evaluation index of the movement path.

[0040] Calculate the variance of each pheromone component, perform multidimensional convergence determination based on the relative rate of change of the variance, and output the globally optimal cleaning path when the multidimensional convergence determination result is less than a preset convergence threshold.

[0041] In one alternative implementation,

[0042] The multi-robot collaborative cleaning control steps include:

[0043] The dirt density distribution and area of ​​the cleaning area are obtained, the workload of the sub-area is calculated, and the cleaning area is divided into multiple sub-areas and assigned to multiple cleaning robots based on the balance of the workload of the sub-areas.

[0044] The cleaning trajectory in each sub-region is divided into high-density trajectory segments and low-density trajectory segments according to the dirt density threshold. A first sliding surface is constructed by weighted combination of velocity error and position error for the high-density trajectory segments, and a second sliding surface is constructed by weighted combination of position error integral term and position error for the low-density trajectory segments.

[0045] The equivalent control quantity is calculated based on the state quantities of the first and second sliding surfaces, and the equivalent control quantity is used to compensate for the deterministic dynamic characteristics during the cleaning process; the adaptive switching gain is calculated based on the dirt density, cleaning resistance and the state quantities of the sliding surfaces, and the adaptive switching gain is used to suppress uncertain disturbances during the cleaning process; the equivalent control quantity and the sign function of the adaptive switching gain are combined to generate a piecewise adaptive sliding mode control input.

[0046] Calculate the percentage of cleaned area and distance from the target point for each cleaning robot to determine the obstacle avoidance priority; when the distance between adjacent cleaning robots is less than the safe distance threshold, the segmented adaptive sliding mode control input is corrected according to the obstacle avoidance priority, and the correction amount is proportional to the difference in obstacle avoidance priority between adjacent robots and the distance gradient.

[0047] A second aspect of the present invention,

[0048] Provides an environment-aware adaptive cleaning path planning system for photovoltaic panels, including:

[0049] The first unit is used to acquire historical weather data and photovoltaic panel status monitoring data of the area where the photovoltaic panel is located, and calculate the cleanliness value of the photovoltaic panel; when the cleanliness value is lower than the cleaning threshold dynamically calculated based on power generation loss and cleaning cost, a cleaning trigger command is generated.

[0050] The second unit is used to receive the cleaning trigger command and execute the cleaning path optimization steps, including: setting the optimization objectives as maximizing cleaning efficiency, minimizing energy consumption, and optimizing time cost; dynamically adjusting the weights based on environmental impact and cleaning energy efficiency ratio; optimizing the population using a genetic algorithm based on dirt density encoding; mapping the optimized population to the initial distribution of cleaning efficiency pheromones, energy consumption pheromones, and time cost pheromones; obtaining the global path through iterative calculation using a multi-pheromone ant colony algorithm; and performing local optimization using an environmentally adaptive multi-neighborhood simulated annealing algorithm.

[0051] The third unit is used to execute multi-robot collaborative cleaning control steps based on the optimized cleaning path, including: dividing the area into sub-regions based on dirt density and workload balance; using segmented adaptive sliding mode variable structure control for trajectory tracking; and determining obstacle avoidance priority based on the percentage of cleaned area and target distance for multi-robot collaborative control.

[0052] A third aspect of the present invention,

[0053] An electronic device is provided, comprising:

[0054] processor;

[0055] Memory used to store processor-executable instructions;

[0056] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0057] Fourth aspect of the embodiments of the present invention,

[0058] A computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0059] The beneficial effects of this application are as follows:

[0060] By acquiring historical weather data and status monitoring data of the area where the photovoltaic panels are located, and dynamically calculating the cleanliness value and cleaning threshold, intelligent decision-making on the timing of cleaning is achieved, avoiding unnecessary cleaning operations. This reduces maintenance costs while ensuring power generation efficiency and improves the overall economic benefits of the photovoltaic system.

[0061] A path planning method integrating multi-objective optimization strategy and multi-algorithm fusion is adopted. Cleaning efficiency, energy consumption and time cost are taken as optimization objectives, and the weights are dynamically adjusted based on environmental impact and cleaning energy efficiency ratio. Through the synergistic effect of genetic algorithm, multi-pheromone ant colony algorithm and environmental adaptive multi-neighborhood simulated annealing algorithm, global optimization and local fine adjustment of cleaning path are achieved, which significantly improves the efficiency and quality of cleaning operation.

[0062] A multi-robot collaborative cleaning control mechanism is introduced, which divides the system into sub-regions based on dirt density and workload balance. It adopts segmented adaptive sliding mode variable structure control technology to achieve precise trajectory tracking. By using an obstacle avoidance priority strategy based on the percentage of cleaned area and target distance, the conflict problem in multi-robot collaborative operation is solved, which greatly improves the system's cleaning efficiency and robustness and adapts to the photovoltaic panel cleaning needs under various complex environmental conditions. Attached Figure Description

[0063] Figure 1 This is a flowchart illustrating the adaptive cleaning path planning method for photovoltaic panels based on environmental perception, as described in an embodiment of the present invention.

[0064] Figure 2 A comparison chart showing the change in cleaning path optimization rate with the number of iterations;

[0065] Figure 3 A comparison chart showing the performance retention rate of different algorithms under different environmental conditions;

[0066] Figure 4 This is a simulation diagram of the cleaning path of the multi-pheromone ant colony algorithm based on grid energy efficiency gradient of the present invention;

[0067] Figure 5 This is a diagram illustrating the convergence process of the multi-pheromone ant colony algorithm of this invention. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0069] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0070] Figure 1This is a flowchart illustrating the adaptive cleaning path planning method for photovoltaic panels based on environmental perception, as described in an embodiment of the present invention. Figure 1 As shown, the method includes:

[0071] Historical weather data and photovoltaic panel status monitoring data of the area where the photovoltaic panel is located are obtained, and the cleanliness value of the photovoltaic panel is calculated; when the cleanliness value is lower than the cleaning threshold dynamically calculated based on power generation loss and cleaning cost, a cleaning trigger command is generated.

[0072] Upon receiving the cleaning trigger command, the cleaning path optimization steps are executed, including: setting the optimization objectives as maximizing cleaning efficiency, minimizing energy consumption, and optimizing time cost; dynamically adjusting the weights based on environmental impact and cleaning energy efficiency ratio; optimizing the population using a genetic algorithm based on dirt density encoding; mapping the optimized population to the initial distribution of cleaning efficiency pheromones, energy consumption pheromones, and time cost pheromones; obtaining the global path through iterative calculation using a multi-pheromone ant colony algorithm; and performing local optimization using an environmentally adaptive multi-neighborhood simulated annealing algorithm.

[0073] Based on the optimized cleaning path, multi-robot collaborative cleaning control steps are executed, including: dividing the area into sub-regions based on dirt density and workload balance; using segmented adaptive sliding mode variable structure control for trajectory tracking; determining obstacle avoidance priority based on the percentage of cleaned area and target distance, and performing multi-robot collaborative control.

[0074] In one optional implementation, acquiring historical weather data of the area where the photovoltaic panel is located and status monitoring data of the photovoltaic panel, and calculating the cleanliness value of the photovoltaic panel includes:

[0075] Historical weather data is processed using a sliding time window method. A time decay factor is introduced into the data within each time window for weighted calculation to obtain the historical weather impact value.

[0076] The historical weather impact values ​​and status monitoring data are multiplied by the corresponding weight coefficients and summed to obtain the photovoltaic panel cleanliness value. The initial values ​​of each weight coefficient are determined by fuzzy hierarchical analysis, and the gradient direction of the weights is dynamically adjusted based on the cleanliness value during the calculation process.

[0077] Establish a historical cleaning record database to record the cleanliness value before each cleaning, cleaning cost, and actual power generation loss during the corresponding period. Combine seasonal and environmental factors to calculate the power generation loss caused by the reduction in unit cleanliness.

[0078] A cleaning benefit evaluation function is established based on the unit cleaning cost and the power generation loss caused by the reduction in unit cleanliness. The corresponding cleaning benefit is calculated by traversing different cleanliness values, and the cleanliness value that maximizes the cleaning benefit is determined as the cleaning threshold for the current period. The cleaning benefit evaluation function also considers the cost of cleaning resource scheduling and the time window constraint of cleaning operation.

[0079] For example, the system acquires historical weather data and photovoltaic panel status monitoring data for the area where the photovoltaic panels are located. Historical weather data includes rainfall, snowfall, wind speed, wind direction, humidity, and temperature; status monitoring data includes photovoltaic panel surface temperature, power generation, current, voltage, and irradiance.

[0080] For processing historical weather data, a sliding time window method is used. Specifically, a time window is set, such as 14 days, and weather data within the window is processed. A time decay factor is introduced to weight data at different time points, so that data closer to the current time has a greater weight. The impact value of different weather factors is calculated separately. For example, for rainfall, the daily rainfall data is first multiplied by the corresponding time decay factor, and then summed to obtain the weighted rainfall. Similarly, weighted wind speed, weighted humidity, etc., are calculated. Finally, the weighted values ​​of all weather factors are combined to obtain the historical weather impact value.

[0081] Historical weather impact values ​​and condition monitoring data are combined to calculate the cleanliness value of photovoltaic panels. Weighting coefficients are set for each indicator, with initial weights determined using fuzzy hierarchical analysis. For example, historical weather impact values ​​have a weight of 0.4, power generation efficiency a weight of 0.3, surface temperature a weight of 0.2, and current-voltage characteristics a weight of 0.1. Each indicator value is multiplied by its corresponding weight and summed to obtain the photovoltaic panel cleanliness value, ranging from 0 to 100, with lower values ​​indicating more severe pollution. During the calculation process, the system dynamically adjusts the gradient direction of the weights based on the cleanliness value. Specifically, by comparing the calculated cleanliness value with the cleanliness value measured before actual cleaning, the weights of each indicator are adjusted to make the calculation results more accurate. The adjustment magnitude is proportional to the size of the difference, but upper and lower limits are set to ensure stability.

[0082] For example, a cleanliness value calculated in a certain instance might be 75, while the actual measurement before cleaning might be 70, a difference of 5 units. The system will appropriately increase the weight of indicators that indicate more severe pollution, such as adjusting the weight of historical weather impact values ​​from 0.4 to 0.42, while reducing the weight of other indicators to ensure that the total weight is 1.

[0083] The system establishes a historical cleaning record database, recording the cleanliness level before each cleaning, the cleaning cost, and the actual power generation loss for the corresponding period. Cleaning costs include labor costs, water resource costs, and equipment depreciation costs; power generation loss is calculated by comparing the power generation efficiency before and after cleaning.

[0084] Combining seasonal and environmental factors, the power generation loss caused by a reduction in unit cleanliness is calculated. The seasonal factor considers the differences in solar irradiance intensity in different seasons, for example, the coefficient is 1.2 in summer and 0.8 in winter; the environmental factor considers the differences in the impact of pollutants on power generation efficiency under different environments, for example, the coefficient is 1.3 in industrial areas and 0.9 in rural areas.

[0085] Historical data analysis reveals the power generation loss caused by a decrease in unit cleanliness (e.g., from 100 to 99). For example, during the summer, when the cleanliness level of a photovoltaic power station decreased from 95 to 85, the average daily power generation dropped from 500 kWh to 470 kWh. Therefore, the power generation loss caused by the decrease in unit cleanliness is (500-470) / 10 = 3 kWh / unit.

[0086] A cleaning benefit evaluation function is established based on the unit cleaning cost and the power generation loss caused by the reduction in unit cleanliness. The cleaning benefit equals the increased power generation revenue after cleaning minus the cleaning cost. The increased power generation revenue is calculated by multiplying the increase in cleanliness by the increase in power generation per unit of cleanliness by the electricity price.

[0087] The evaluation of cleaning benefits also considers the cost of cleaning resource scheduling and the time window constraint for cleaning operations. The cost of cleaning resource scheduling is related to the distance the cleaning equipment needs to be moved and the difficulty of personnel deployment; the time window constraint considers factors such as weather conditions and power grid dispatch requirements.

[0088] By iterating through different cleanliness values ​​(e.g., from 50 to 90, with a step size of 1), the benefits of cleaning at each cleanliness value are calculated. For example, when the cleanliness is 80, the cleaning cost is 200 yuan. After cleaning, the cleanliness increases to 98, increasing power generation by 18 × 3 = 54 kWh. With an electricity price of 0.8 yuan / kWh, the increased revenue is 54 × 0.8 = 43.2 yuan. The cleaning benefit is 43.2 - 200 = -156.8 yuan, which is negative, indicating that cleaning is uneconomical at this point.

[0089] The system identifies the cleanliness value that maximizes cleaning efficiency and sets it as the cleaning threshold for the current period. When the actual cleanliness level falls below this threshold, the system will recommend cleaning. For example, if calculations show that a cleanliness level of 65 maximizes the cleaning efficiency to 300 yuan, then 65 is set as the cleaning threshold.

[0090] This invention, by comprehensively considering the impact of weather, the degree of pollution, cleaning costs, and power generation benefits, achieves accurate assessment of photovoltaic panel cleanliness and intelligent cleaning decisions, effectively improving the operation and maintenance efficiency and economic benefits of photovoltaic power generation systems.

[0091] In one optional implementation, dynamically adjusting the weights based on environmental impact and cleaning energy efficiency ratio includes:

[0092] An environmental impact function is established to characterize the effects of temperature, humidity, wind speed, light intensity, and precipitation probability on the cleaning effect. The temperature parameter is calculated using an exponential function to determine the degree of influence of temperature deviation from the optimal cleaning temperature, and the humidity parameter is calculated using an S-shaped response curve to determine the degree of influence of humidity exceeding a critical value. A coupling coefficient between the temperature and humidity parameters is constructed based on the environmental impact function.

[0093] The grid cells are divided based on the density of dirt distribution on the photovoltaic panel surface, and initial weights are set for cleaning efficiency, energy consumption, and time cost.

[0094] An environmental adjustment coefficient is calculated based on the environmental impact function and the coupling coefficient, and the environmental adjustment coefficient is positively correlated with the environmental sensitivity of the grid cell; an energy efficiency ratio adjustment coefficient is calculated based on the cleaning energy efficiency ratio of the grid cell, and the energy efficiency ratio adjustment coefficient is a power function relationship between the current cleaning energy efficiency ratio and the maximum cleaning energy efficiency ratio;

[0095] The product of the environmental adjustment coefficient and the energy efficiency ratio adjustment coefficient is used as the weight adjustment amount, so that the cleaning efficiency weight is negatively correlated with the weight adjustment amount, and the energy consumption weight and the time cost weight are positively correlated with the weight adjustment amount, and the initial weights are dynamically updated.

[0096] For example, an environmental impact function is established to characterize the influence of various environmental parameters on the cleaning effect. The temperature parameter's influence is calculated using an exponential function. Specifically, when the actual temperature is 25℃ and the optimal cleaning temperature is 20℃, the temperature deviation is 5℃, and the temperature influence coefficient is 0.78 after substituting into the exponential function. The humidity parameter's influence is calculated using an S-shaped response curve. For example, when the actual humidity is 75% and the critical humidity is 60%, the humidity influence coefficient is 0.65 after substituting into the S-shaped response function. The wind speed parameter is processed using a piecewise function. The influence coefficient is 0.9 when the wind speed is below 3m / s, linearly decreases to 0.5 when the wind speed is between 3-8m / s, and is 0.3 when the wind speed exceeds 8m / s. The light intensity parameter is mapped using a logarithmic function. For example, when the light intensity is 800W / m²... 2 When the probability of precipitation is below 20%, the corresponding influence coefficient is 0.85. The precipitation probability parameter uses a threshold method: the influence coefficient is 0.95 when the probability of precipitation is below 20%, 0.6 when the probability is between 20% and 50%, and 0.2 when the probability exceeds 50%. The coupling coefficient between temperature and humidity parameters is calculated by weighting the influence coefficients of temperature and humidity, with weights of 0.6 and 0.4 respectively. For example, when the temperature influence coefficient is 0.78 and the humidity influence coefficient is 0.65, the calculated coupling coefficient is 0.726.

[0097] The grid cells were divided based on the surface dirt distribution density of the photovoltaic panel. A standard photovoltaic panel was divided into a 6×8 grid, resulting in 48 grid cells. Initial weights were set for cleaning efficiency, energy consumption, and time cost for each grid cell, with initial values ​​of 0.5, 0.3, and 0.2, respectively.

[0098] The environmental adjustment coefficient is calculated based on the environmental impact function and coupling coefficient. The environmental adjustment coefficient is positively correlated with the environmental sensitivity of the grid cell, which is determined by the type and degree of contamination. When the environmental sensitivity of a grid cell is 0.8, the calculated environmental impact function result is 0.7, and the coupling coefficient is 0.726, the calculated environmental adjustment coefficient is 0.8 × 0.7 × 0.726 = 0.406. The energy efficiency ratio adjustment coefficient is calculated based on the cleaning energy efficiency ratio of the grid cell. The cleaning energy efficiency ratio is defined as the ratio of cleaning effectiveness to energy consumption, obtained through real-time monitoring data. For example, if the current cleaning energy efficiency ratio of a grid cell is 2.5, the maximum cleaning energy efficiency ratio in the system's historical records is 4.0, and the power function exponent is set to 0.5, then the energy efficiency ratio adjustment coefficient is (2.5 / 4.0). 0.5 =0.791.

[0099] The product of the environmental adjustment coefficient and the energy efficiency ratio adjustment coefficient is used as the weight adjustment amount. For example, when the environmental adjustment coefficient is 0.406 and the energy efficiency ratio adjustment coefficient is 0.791, the weight adjustment amount is 0.406 × 0.791 = 0.321. The initial weights are dynamically updated so that the cleaning efficiency weight is negatively correlated with the weight adjustment amount, and the energy consumption weight and time cost weight are positively correlated with the weight adjustment amount. In specific implementation, a linear adjustment formula is used: Cleaning efficiency weight = Initial cleaning efficiency weight - Weight adjustment amount × Adjustment factor; Energy consumption weight = Initial energy consumption weight + Weight adjustment amount × Energy adjustment factor; Time cost weight = Initial time cost weight + Weight adjustment amount × Time adjustment factor. The adjustment factors, energy adjustment factor, and time adjustment factor are set to 0.5, 0.3, and 0.2, respectively.

[0100] Using the above example, the adjusted cleaning efficiency weight is calculated as follows: 0.5 - 0.321 × 0.5 = 0.3395; the adjusted energy consumption weight is calculated as follows: 0.3 + 0.321 × 0.3 = 0.3963; and the adjusted time cost weight is calculated as follows: 0.2 + 0.321 × 0.2 = 0.2642.

[0101] In practical applications, the system collects environmental data and updates the weights at fixed intervals. For example, during a complete cleaning cycle, the system recorded the following data: initial temperature of 22℃, humidity of 65%, wind speed of 2m / s, and light intensity of 750W / m². 2The probability of precipitation is 15%; during the mid-term, the temperature rises to 28℃, the humidity decreases to 55%, the wind speed increases to 4m / s, and the solar radiation intensity increases to 850W / m². 2 The probability of precipitation rises to 25%; the ending temperature is 30℃, humidity is 50%, wind speed is 5m / s, and solar irradiance is 900W / m². 2 The probability of precipitation is 30%.

[0102] Based on these environmental changes, the system calculates the environmental adjustment coefficient and energy efficiency ratio adjustment coefficient for each of the 48 grid cells, and then updates the weights. For example, for grid cell A containing water-soluble dirt, the initial environmental adjustment coefficient is calculated to be 0.406, the energy efficiency ratio adjustment coefficient is 0.791, and the weight adjustment is 0.321, resulting in an adjusted weight of [0.3395, 0.3963, 0.2642]. In the middle stage, the environmental adjustment coefficient becomes 0.352, the energy efficiency ratio adjustment coefficient becomes 0.745, and the weight adjustment is 0.262, resulting in an adjusted weight of [0.369, 0.379, 0.252]. At the end, the environmental adjustment coefficient further decreases to 0.301, the energy efficiency ratio adjustment coefficient becomes 0.712, and the weight adjustment is 0.214, resulting in an adjusted weight of [0.393, 0.364, 0.243].

[0103] Existing photovoltaic panel cleaning systems do not adequately consider the interactive effects between environmental parameters, and weight adjustments lack adaptability to the dynamic characteristics of the cleaning process, failing to establish a systematic correlation between environmental impact and cleaning energy efficiency. This invention innovatively proposes a multi-environmental parameter coupling mechanism, employing different function models to characterize the impact of each environmental parameter, constructing a temperature and humidity coupling coefficient to reflect the interaction, and achieving a comprehensive quantitative assessment of environmental impact. It designs a gridded cleaning energy efficiency assessment method, introducing the concept of grid unit environmental sensitivity, and establishing a power function mapping relationship for the cleaning energy efficiency ratio to achieve refined local cleaning effect assessment. It proposes a two-factor dynamic weight adjustment strategy, with the environmental adjustment coefficient and the energy efficiency ratio adjustment coefficient acting together, and a differentiated weight update mechanism to achieve adaptive optimization of the cleaning process. In terms of cleaning quality, this invention improves cleaning uniformity by 15%-25% under different environmental conditions; in terms of energy utilization efficiency, it reduces energy consumption by 20%-30% for the same cleaning effect; in terms of time efficiency, it reduces total cleaning time by 25%-35%; and in terms of adaptability, it shortens the response time to environmental changes by 40%-50%.

[0104] In one optional implementation, a genetic algorithm based on dirt density encoding is used to optimize the population. The optimized population is mapped to the initial distribution of cleaning efficiency pheromones, energy consumption pheromones, and time cost pheromones. The global path is obtained through iterative calculation using a multi-pheromone ant colony algorithm, and local optimization is performed using an environment-adaptive multi-neighborhood simulated annealing algorithm, including:

[0105] The initial population is initialized and constructed using a genetic algorithm, which employs an encoding strategy based on dirt density weights and an environment-aware adaptive crossover and mutation mechanism.

[0106] The optimization results of the genetic algorithm are used as the initial pheromone distribution of the ant colony algorithm. A multi-pheromone ant colony algorithm based on grid energy efficiency gradient is used to generate the globally optimal cleaning path. The multi-pheromone includes cleaning efficiency pheromone, energy consumption pheromone and time cost pheromone.

[0107] The globally optimal cleaning path is locally optimized using an environment-adaptive multi-neighborhood simulated annealing algorithm. The multi-neighborhood includes reverse neighborhood, exchange neighborhood, insertion neighborhood, and region reconstruction neighborhood. Local path optimization is performed through an environment-aware temperature adjustment mechanism and a dynamic acceptance criterion strategy.

[0108] For example, a genetic algorithm is used for initialization and to construct the initial population. In a practical application scenario, assuming the cleaning area is a rectangular area of ​​100m × 100m, it is divided into a 10 × 10 grid, with each grid being 10m × 10m in size. Each grid has a different dirt density value, ranging from 0 to 10, where 0 represents no dirt and 10 represents the highest level of dirt.

[0109] In the genetic algorithm, an encoding strategy based on dirt density weights is employed. Each chromosome represents a cleaning path with a length equal to the total number of grid cells, i.e., 100. Each gene in the chromosome represents the visiting order of a grid cell. The initial population size is set to 50, and 50 distinct paths are randomly generated as the initial population.

[0110] The environmental perception adaptive crossover and mutation mechanism dynamically adjusts the crossover and mutation probabilities based on the dirt density. When the cleaning robot is in a high dirt density area, the crossover probability is set to 0.85 and the mutation probability is set to 0.15; when it is in a medium dirt density area, the crossover probability is set to 0.75 and the mutation probability is set to 0.2; and when it is in a low dirt density area, the crossover probability is set to 0.65 and the mutation probability is set to 0.25.

[0111] In the selection operation, a roulette wheel selection algorithm is used, and the fitness function comprehensively considers path length, dirt removal efficiency, and energy consumption. When calculating the fitness value, the weight of path length is 0.3, the weight of dirt removal efficiency is 0.4, and the weight of energy consumption is 0.3.

[0112] Crossover is performed sequentially, randomly selecting a segment from the parent chromosome for exchange, ensuring that the offspring chromosome does not contain duplicate sequences. Mutation is performed using inversion mutation, randomly selecting two locations on the chromosome and reversing the gene sequence between them.

[0113] After 100 iterations of the genetic algorithm, the 10 individuals with the highest fitness are selected as the optimization results and used to initialize the pheromone distribution of the ant colony algorithm.

[0114] The optimization results of the genetic algorithm are used as the initial pheromone distribution for the ant colony algorithm. A multi-pheromone ant colony algorithm based on grid energy efficiency gradient is used to generate the globally optimal cleaning path.

[0115] The pheromones include cleaning efficiency pheromones, energy consumption pheromones, and time cost pheromones. The cleaning efficiency pheromone is directly proportional to the mesh dirt density; a mesh with a dirt density of 10 has an initial pheromone value of 1.0, a mesh with a dirt density of 5 has an initial pheromone value of 0.5, and so on. The energy consumption pheromone is related to the mesh terrain complexity; the initial value is 0.2 for flat areas, 0.5 for areas with slight obstacles, and 0.8 for areas with many obstacles. The time cost pheromone is inversely proportional to the distance between meshes; an initial value of 0.9 for adjacent meshes, 0.7 for meshes at a distance of 2, 0.5 for meshes at a distance of 3, and so on.

[0116] When selecting the next grid to visit, ants consider the influence of three pheromones: cleaning efficiency (0.5 weight), energy consumption (0.3 weight), and time cost (0.2 weight). The total number of ants is set to 30, and the number of iterations is 200.

[0117] In each iteration, the pheromone evaporation rate is set to 0.1, meaning 90% of the original pheromone is retained. After an ant completes a full path, it adds pheromone to the grids it traversed, based on the path quality; the amount added is proportional to the path quality. Path quality evaluation metrics include total cleaning efficiency, total energy consumption, and total time cost, with weights of 0.5, 0.3, and 0.2, respectively.

[0118] To avoid the algorithm getting trapped in local optima, the upper limit of pheromone is set to 10.0, and the lower limit is set to 0.1. When the pheromone level of a path exceeds the upper limit, it is restricted to the upper limit value; when it falls below the lower limit, it is set to the lower limit value.

[0119] After 200 iterations, the path with the strongest pheromones was selected as the globally optimal cleaning path. In actual testing, this path was able to cover all dirty areas, with a total path length of 1250m, a cleaning efficiency of 95%, and energy consumption of 75% of the standard value.

[0120] Finally, the environment-adaptive multi-neighborhood simulated annealing algorithm is used to locally optimize the globally optimal cleaning path. The multi-neighborhood includes reverse-order neighborhood, exchange neighborhood, insertion neighborhood, and region reconstruction neighborhood. Reverse-order neighborhood reverses the grid access order between two randomly selected points in the path; exchange neighborhood swaps the access order of two randomly selected grids; insertion neighborhood inserts a randomly selected grid into another position on the path; and region reconstruction neighborhood re-plans the access order within a randomly selected sub-region.

[0121] The environmental sensing temperature regulation mechanism dynamically adjusts the annealing temperature based on the current environmental characteristics. The initial temperature is set to 100. In areas with high dirt density, the temperature drop coefficient is 0.95; in areas with medium dirt density, the temperature drop coefficient is 0.9; and in areas with low dirt density, the temperature drop coefficient is 0.85.

[0122] The dynamic acceptance criterion strategy adjusts the acceptance probability dynamically based on the current iteration number and the quality of the solution. In the early stages of iteration, even if the new solution is slightly worse than the current solution, there is a high probability of acceptance; as the iteration progresses, the acceptance probability gradually decreases, and the algorithm tends to accept the better solution.

[0123] The simulated annealing algorithm iterates 300 times, with a termination temperature set to 0.01. In each iteration, a neighborhood operation is randomly selected to generate a new solution, and the acceptance of the new solution is determined based on an acceptance criterion.

[0124] After local optimization, the cleaning path was further optimized, the total path length was reduced to 1180m, the cleaning efficiency was increased to 97%, and the energy consumption was reduced to 70% of the standard value, which is significantly better than the global path optimization.

[0125] This invention achieves efficient and energy-saving intelligent cleaning path planning through a three-stage fusion optimization of genetic algorithm, multi-pheromone ant colony algorithm, and environment-adaptive multi-neighborhood simulated annealing algorithm. The contamination density encoding strategy and environmental perception adaptive mechanism of the genetic algorithm provide a high-quality initial solution for path optimization; the multi-pheromone ant colony algorithm effectively balances cleaning quality and resource consumption through the synergistic effect of three pheromones: cleaning efficiency, energy consumption, and time cost; and the environment-adaptive multi-neighborhood simulated annealing algorithm further realizes local fine-grained optimization of the path.

[0126] In one optional implementation, the genetic algorithm employs an encoding strategy based on dirt density weights and an environment-aware adaptive crossover and mutation mechanism, including:

[0127] The characteristics of various types of dirt on the surface of the photovoltaic panel are obtained, the surface of the photovoltaic panel is divided into grid cells, and the dirt density of each grid cell is calculated.

[0128] The chromosome of the genetic algorithm is constructed as a multi-level coding structure of spatial sequence coding, resource allocation coding and density correlation coding, where spatial sequence coding represents the cleaning path, resource allocation coding represents the cleaning parameters, and density correlation coding represents the dirt handling strategy.

[0129] A density weight mapping function is constructed based on the dirt density. The density weight mapping function simultaneously considers the dirt density of the target grid cell and the dirt density distribution of its adjacent grid cells. The population is initialized according to the density weight mapping function.

[0130] Environmental parameters are acquired and environmental impact factors are calculated. Based on the environmental impact factors and the density weight mapping function, the crossover probability and mutation probability of the genetic algorithm are dynamically adjusted. The cleaning path sequence is optimized through crossover operation, and the resource allocation strategy is adjusted through mutation operation.

[0131] The spatial sequence encoding, resource allocation encoding, and density correlation encoding of the best individual in the optimized population are mapped to the initial pheromone distribution matrix of the ant colony algorithm.

[0132] For example, multiple types of dirt features on the surface of a photovoltaic panel are acquired. Image data of the photovoltaic panel surface is obtained through an image acquisition device, and image processing technology is used to identify different types of dirt, such as dust, bird droppings, and leaves. The acquired image data is preprocessed, including noise reduction and contrast enhancement, to improve the accuracy of dirt identification. For example, for a 1.5m × 1m photovoltaic panel, a high-definition image of 2000 × 1500 pixels can be acquired, and different types of dirt areas can be identified through HSV color space conversion and threshold segmentation. The surface of the photovoltaic panel is divided into grid cells, and the dirt density of each grid cell is calculated. The surface of the photovoltaic panel is uniformly divided into m × n grid cells, such as a 10 × 8 grid. For each grid cell, the ratio of the number of dirty pixels to the total number of pixels in the grid cell is used as the dirt density of that grid cell. For example, for a grid cell numbered (3, 4), if it contains 5000 pixels, of which 2000 are dirty pixels, then the dirt density of that grid cell is 0.4.

[0133] The multi-level chromosome coding structure of the genetic algorithm consists of three parts: spatial sequence coding, resource allocation coding, and density-association coding. Spatial sequence coding represents the cleaning path, using integer coding with a length equal to the total number of grid cells. Each gene value represents the grid cell number visited. For example, for a 10×8 grid, the spatial sequence coding length is 80, and the code "1-25-36-42..." indicates the cleaning order is grid cells 1, 25, 36, 42, etc. Resource allocation coding represents cleaning parameters, including cleaning pressure and cleaning agent dosage, using real numbers. For example, the code "0.5-0.3-0.8..." indicates that the corresponding grid cell has a cleaning pressure of 0.5 MPa, a cleaning agent dosage of 0.3 ml, and a cleaning time of 0.8 s. Density-association coding represents the contamination treatment strategy, using binary coding to indicate whether special treatment is required. For example, the code "1-0-1..." indicates that the first and third cells in the corresponding grid cell require special treatment, while the second cell is treated normally.

[0134] A density weighting mapping function is constructed based on dirt density. This function considers not only the dirt density of the target grid cell but also the dirt density distribution of its neighboring grid cells. For any grid cell, the density weight value is calculated as follows: first, the dirt density value of the grid cell is calculated; then, the average dirt density value of its eight neighboring grid cells is calculated; finally, the two density values ​​are weighted and summed to obtain the final density weight value. The weighting coefficients can be adjusted according to the actual situation. For example, the weight of the target grid cell's dirt density is 0.7, and the weight of the average dirt density of neighboring cells is 0.3. For edge grid cells, only existing neighboring cells are considered. For example, for the grid cell (5, 6) with a dirt density of 0.6, the average dirt density of its neighboring cells is 0.4, so its density weight value is 0.6 × 0.7 + 0.4 × 0.3 = 0.54.

[0135] The population is initialized using a density-weighted mapping function. The population size is set to 100. For each individual, the spatial sequence encoding prioritizes grid cells with high density weights to form the initial cleaning path. The resource allocation encoding sets initial values ​​based on the dirt density of the corresponding grid cell; higher dirt density results in more cleaning resources being allocated. The density association encoding is set based on the dirt type; difficult-to-remove dirt types are set to 1, indicating that special handling is required.

[0136] Obtain environmental parameters and calculate environmental impact factors. Environmental parameters include ambient temperature, humidity, and light intensity. For example, when the ambient temperature is 30℃, the humidity is 60%, and the light intensity is 800W / m²... 2At this time, the environmental impact factor can be set to 0.85. The higher the ambient temperature, the easier it is for dirt to dry, increasing the difficulty of cleaning; the higher the humidity, the easier it is for dirt to soften, reducing the difficulty of cleaning; the greater the light intensity, the shorter the cleaning time window.

[0137] The crossover and mutation probabilities of the genetic algorithm are dynamically adjusted based on the environmental impact factor and density weight mapping function. The initial crossover probability is set to 0.8, and the mutation probability to 0.1. The crossover probability is adjusted according to the environmental impact factor; a higher environmental impact factor results in a higher crossover probability, enhancing the algorithm's global search capability. The mutation probability is adjusted according to the density weight mapping function; regions with higher density weight values ​​have lower mutation probabilities, preserving superior genes. For example, when the environmental impact factor is 0.85, the adjusted crossover probability is 0.8 × 0.85 = 0.68; when the density weight of a certain grid cell is 0.54, the mutation probability at that location is 0.1 × (1 - 0.54) = 0.046.

[0138] The cleaned path sequence is optimized through crossover operations. The Partial Matching Crossover (PMX) method is used to crossover spatial sequence codes, ensuring that the crossover codes remain valid paths. For example, given the spatial sequence codes "1-3-5-7-9" for parent individual A and "2-4-6-8-10" for parent individual B, the crossover points are selected at positions 2 and 4, resulting in the spatial sequence codes "1-4-6-7-9" for offspring individual C and "2-3-5-8-10" for offspring individual D.

[0139] Resource allocation strategies are adjusted through mutation operations. Gaussian mutation is performed on the resource allocation code, with the mutation magnitude proportional to the dirt density of the corresponding grid cell. For example, for the resource allocation code "0.5-0.3-0.8", if the dirt density of the corresponding grid cell is 0.6, the mutation may result in "0.55-0.3-0.8", increasing the cleaning pressure to cope with the higher dirt density.

[0140] The spatial sequence encoding, resource allocation encoding, and density association encoding of the optimal individual in the optimized population are mapped to the initial pheromone distribution matrix for the ant colony algorithm. For adjacent grid cell pairs in the spatial sequence encoding, initial pheromone values ​​are set at the corresponding positions in the pheromone matrix, with the values ​​proportional to the density weights of these two grid cells. For the resource allocation encoding and density association encoding, they are transformed into decision probability adjustment factors for ants in the ant colony algorithm. For example, for the spatial sequence encoding of the optimal individual "1-25-36-42...", higher initial pheromone values ​​are set at positions (1, 25), (25, 36), and (36, 42) in the pheromone matrix to guide ants to explore along the optimized path.

[0141] Figure 2 The graph shows a comparison of the cleaning path optimization rate as a function of the number of iterations. Figure 2 As shown, the algorithm of this invention (solid black line in the figure) achieved a path optimization rate of over 20% in just 40 iterations, while the traditional single-encoding algorithm (dashed line in the figure) required more than 80 iterations to reach the same level. The performance difference between the two algorithms becomes more pronounced as the number of iterations increases; after 100 iterations, the path optimization rate of the algorithm of this invention is higher than the other two algorithms. More notably, the algorithm of this invention exhibits a faster convergence speed and a steeper curve slope in the early stages of iteration. This fully demonstrates that the encoding strategy based on dirt density weights can effectively improve the search efficiency and solution quality of the algorithm, providing a superior solution for photovoltaic panel cleaning path planning.

[0142] like Figure 3 The performance retention rate comparison chart of different algorithms under different environmental conditions is shown. The horizontal axis represents the magnitude of environmental condition changes (from 10% to 90%), and the vertical axis represents the algorithm performance retention rate (from 50% to 100%). The environmental adaptive algorithm of this invention (solid black line and solid dots) exhibits significant robustness advantages when environmental conditions change drastically. When the environmental change reaches 90%, the algorithm of this invention can still maintain approximately 78% of its performance, while the performance of the traditional fixed-parameter algorithm (dashed line and hollow square) drops to approximately 58%. The adaptive mechanism of this invention can effectively cope with environmental changes, providing a reliable guarantee for the stable operation of the photovoltaic panel cleaning system in complex and variable environments.

[0143] This invention combines a multi-level coding structure with an environmental perception and adaptive mechanism to achieve intelligent optimization of photovoltaic panel cleaning paths and resource allocation, providing an effective solution for improving photovoltaic system efficiency and reducing maintenance costs.

[0144] In one optional implementation, the optimization result of the genetic algorithm is used as the initial pheromone distribution for the ant colony algorithm. A multi-pheromone ant colony algorithm based on grid energy efficiency gradient is then used to generate the globally optimal cleaning path. The multi-pheromone includes cleaning efficiency pheromone, energy consumption pheromone, and time cost pheromone.

[0145] The optimization result of the genetic algorithm is mapped to a multidimensional pheromone initial distribution, which includes cleaning efficiency pheromone, energy consumption pheromone and time cost pheromone. Each pheromone component is obtained by weighted combination of the sequence information corresponding to the optimal individual and the grid dirt density.

[0146] The grid energy efficiency ratio is calculated based on the ratio of cleaning revenue to energy consumption of the grid cells, and the grid energy efficiency gradient vector is constructed based on the partial derivatives of the grid energy efficiency ratio in the horizontal and vertical directions.

[0147] The heuristic information of each pheromone component is weighted and multiplied to obtain the comprehensive heuristic information. Based on the comprehensive heuristic information, the initial distribution of multidimensional pheromones, and the grid energy efficiency gradient vector, the state transition probability matrix is ​​calculated.

[0148] The ant's movement path is determined based on the state transition probability matrix. Local pheromone updates are performed on the grid cells along the movement path, taking into account the influence of the pheromone evaporation coefficient and the grid energy efficiency ratio. The global optimal path is then enhanced with pheromone based on the evaluation index of the movement path.

[0149] Calculate the variance of each pheromone component, perform multidimensional convergence determination based on the relative rate of change of the variance, and output the globally optimal cleaning path when the multidimensional convergence determination result is less than a preset convergence threshold.

[0150] A Cleaning Path Optimization Method Combining Genetic Algorithm and Multi-Pheromone Ant Colony Algorithm

[0151] For example, the optimization results of the genetic algorithm are used as the initial pheromone distribution for the ant colony algorithm, and a multi-pheromone ant colony algorithm based on grid energy efficiency gradient is used to generate the globally optimal cleaning path. The specific implementation process is as follows:

[0152] During the genetic algorithm optimization phase, the cleaning area is divided into grids. For example, a 10m x 10m cleaning area is divided into 100 1m x 1m grid cells. Each grid cell has a dirt density value, a decimal between 0 and 1, with a higher value indicating a higher degree of dirtiness. The initial path is optimized using a genetic algorithm to obtain an optimized path sequence, such as [25, 26, 36, 46, 56, 55, 54, 53, 43, 33, 23, 13], representing the grid cell numbers that the cleaning robot passes through sequentially.

[0153] When mapping the optimization results of the genetic algorithm to a multidimensional pheromone initial distribution, three types of pheromones are considered: cleaning efficiency pheromone, energy consumption pheromone, and time cost pheromone. For the start and end points of adjacent grid cells in the path, the initial pheromone value is calculated by weighting the sequence information corresponding to the optimal individual with the grid dirt density. For example, if the start and end grid cells are adjacent in the optimal path and their dirt densities are 0.8 and 0.7 respectively, then the initial value of the cleaning efficiency pheromone can be set to 0.5 × (0.8 + 0.7) = 0.75; similarly, the initial pheromone values ​​for energy consumption and time cost are calculated.

[0154] The grid energy efficiency ratio is calculated based on the ratio of cleaning revenue to energy consumption per grid cell: cleaning revenue is defined as the dirt density of the grid cell multiplied by the cleaning area, and energy consumption includes both movement energy and cleaning energy consumption. For example, for a grid cell with a dirt density of 0.8, a cleaning area of ​​1 square meter, a cleaning energy consumption of 0.2 kWh, and a movement energy consumption of 0.05 kWh, the grid energy efficiency ratio is (0.8 × 1) / (0.2 + 0.05) = 3.2, representing the cleaning revenue per unit of energy consumption.

[0155] The grid energy efficiency gradient vector is constructed based on the partial derivatives of the grid energy efficiency ratio in the horizontal and vertical directions. Assuming the energy efficiency ratios of two adjacent points in the horizontal direction of an adjacent grid cell are 3.2 and 3.5 respectively, the horizontal partial derivative is 3.5 - 3.2 = 0.3; the energy efficiency ratios of two adjacent points in the vertical direction of an adjacent grid cell are 3.2 and 2.9 respectively, the vertical partial derivative is 2.9 - 3.2 = -0.3. Therefore, the grid energy efficiency gradient vector is constructed as (0.3, -0.3), indicating the direction and magnitude of energy efficiency growth.

[0156] The heuristic information of each pheromone component is weighted and multiplied to obtain the comprehensive heuristic information. The cleaning efficiency heuristic information is directly proportional to the mesh dirt density; the energy consumption heuristic information is inversely proportional to the movement distance and cleaning energy consumption; and the time cost heuristic information is inversely proportional to the movement time and cleaning time. With weighting coefficients of 0.4, 0.3, and 0.3, the comprehensive heuristic information is equal to the weighted product of the individual heuristic information.

[0157] The state transition probability matrix is ​​calculated based on comprehensive heuristic information, the initial distribution of multidimensional pheromones, and the grid energy efficiency gradient vector. The probability of an ant moving from the current grid cell to a candidate grid cell is considered, taking into account the weighted product of the three pheromones, the comprehensive heuristic information, and the grid energy efficiency gradient vector. For example, if the pheromone weights are set to 0.35, 0.35, and 0.3, and the gradient influence factor is 0.2, then the state transition probability is proportional to the product of the pheromone weights, the heuristic information, and the gradient angle factor, where the angle factor is the cosine of the angle between the ant's movement direction and the energy efficiency gradient vector.

[0158] The ant's movement path is determined based on the state transition probability matrix, and local pheromone updates are performed on the grid cells along the path. With the pheromone evaporation coefficient set to 0.1 and the local update intensity factor to 0.05, the local pheromone update formula is: original pheromone value multiplied by (1 - evaporation coefficient) plus the local update intensity factor multiplied by the grid energy efficiency ratio. For example, if the original pheromone value is 0.75 and the grid energy efficiency ratio is 3.2, then the updated pheromone value is (1 - 0.1) × 0.75 + 0.05 × 3.2 = 0.835.

[0159] For the globally optimal path, pheromone enhancement is applied. The global update intensity factor is set to 0.2, and the global pheromone update formula is: original pheromone value multiplied by (1 - volatility coefficient) plus the global update intensity factor multiplied by a constant divided by the total path length. For example, if the globally optimal path length is 25 meters and the constant is 100, the pheromone increment is 0.2 × 100 / 25 = 0.8.

[0160] The variance of each pheromone component is calculated, and multidimensional convergence is determined based on the relative rate of change of the variance. For example, after two consecutive iterations, the variance of the cleaning efficiency pheromone decreases from 0.05 to 0.03, with a relative rate of change of (0.05-0.03) / 0.05 = 0.4; the variance of the energy consumption pheromone decreases from 0.06 to 0.04, with a relative rate of change of 0.33; and the variance of the time cost pheromone decreases from 0.04 to 0.03, with a relative rate of change of 0.25. The average of the three, 0.33, is taken as the multidimensional convergence determination result. If the preset convergence threshold is 0.1, the algorithm continues to iterate; when the multidimensional convergence determination result is less than 0.1, the globally optimal cleaning path is output.

[0161] like Figure 4 The simulation diagram of the cleaning path based on the multi-pheromone ant colony algorithm of the grid energy efficiency gradient of this invention is shown. Different gray levels in the diagram represent the dirt density distribution of each grid cell within a 10×10 meter area. Dark areas represent locations with high dirt density and urgent cleaning needs, while light areas represent locations with lighter dirt levels. Arrows represent the grid energy efficiency gradient vector, indicating the direction of energy efficiency ratio improvement. The solid black line represents the optimal cleaning path generated by the algorithm of this invention, which adopts an optimized serpentine pattern to ensure full coverage while minimizing path length. Simulation results show that the algorithm can rationally plan the path according to the characteristics of dirt distribution. Especially in the heavily soiled area from (3,3) to (5,5) in the central region, the path planning is more compact, reflecting the algorithm's sensitivity to dirt density and rational allocation of cleaning resources. Compared with traditional algorithms, this path reduces the travel distance by approximately 9.1% while increasing the dirt removal rate by 3.5%, effectively solving the multi-objective optimization problem in photovoltaic panel cleaning.

[0162] like Figure 5The convergence process analysis diagram of the multi-pheromone ant colony algorithm of this invention is shown in the figure. The horizontal axis represents the number of algorithm iterations, and the vertical axis represents the variance value of each pheromone. The three curves with different line types represent the variance change trends of the cleaning efficiency pheromone (solid line), energy consumption pheromone (dashed line), and time cost pheromone (dotted line), respectively. The variance values ​​of the three pheromones all show a decreasing trend with iteration, indicating that the ant colony's choice of the optimal path gradually becomes more consistent. The energy consumption pheromone converges the fastest, with the highest initial variance (0.25) but the fastest decrease rate, indicating that energy factors play a dominant role in the optimization process. At the 42nd iteration (the multi-dimensional convergence point marked in the figure), the comprehensive convergence judgment result of the three pheromones is lower than the preset threshold line (0.1), and the algorithm reaches the convergence condition and stops iterating. This multi-dimensional convergence mechanism effectively avoids the local optimum problem caused by premature convergence of a single pheromone.

[0163] This invention combines genetic algorithms and multi-pheromone ant colony algorithms, and uses grid energy efficiency gradients to guide path optimization, reducing travel distance while improving dirt removal rate, effectively solving the multi-objective optimization problem of cleaning tasks.

[0164] In one optional implementation, the multi-robot collaborative cleaning control steps include:

[0165] The dirt density distribution and area of ​​the cleaning area are obtained, the workload of the sub-area is calculated, and the cleaning area is divided into multiple sub-areas and assigned to multiple cleaning robots based on the balance of the workload of the sub-areas.

[0166] The cleaning trajectory in each sub-region is divided into high-density trajectory segments and low-density trajectory segments according to the dirt density threshold. A first sliding surface is constructed by weighted combination of velocity error and position error for the high-density trajectory segments, and a second sliding surface is constructed by weighted combination of position error integral term and position error for the low-density trajectory segments.

[0167] The equivalent control quantity is calculated based on the state quantities of the first and second sliding surfaces, and the equivalent control quantity is used to compensate for the deterministic dynamic characteristics during the cleaning process; the adaptive switching gain is calculated based on the dirt density, cleaning resistance and the state quantities of the sliding surfaces, and the adaptive switching gain is used to suppress uncertain disturbances during the cleaning process; the equivalent control quantity and the sign function of the adaptive switching gain are combined to generate a piecewise adaptive sliding mode control input.

[0168] Calculate the percentage of cleaned area and distance from the target point for each cleaning robot to determine the obstacle avoidance priority; when the distance between adjacent cleaning robots is less than the safe distance threshold, the segmented adaptive sliding mode control input is corrected according to the obstacle avoidance priority, and the correction amount is proportional to the difference in obstacle avoidance priority between adjacent robots and the distance gradient.

[0169] For example, the dirt density distribution and area of ​​the cleaning area are obtained. The dirt density distribution can be obtained by capturing images through a camera mounted on the robot and calculating them using an image processing algorithm. For example, the cleaning area is divided into a 10×10 grid, with each grid having a dirt density value ranging from 0 to 10, where 0 represents completely clean and 10 represents extremely dirty. The area is obtained through measurement, such as an office floor area of ​​100 square meters. Based on the obtained dirt density distribution, the workload of each sub-area is calculated. The workload calculation formula is the product of the sub-area area and the average dirt density of that area. For example, if a 5-square-meter sub-area has an average dirt density of 7, then the workload for that area is 35 work units. To achieve workload balance, an iterative bisection method is used to divide the cleaning area into multiple sub-areas. The specific steps are as follows: First, the area is divided into two parts along the dimension with the largest area, and the difference in workload between the two parts is calculated; if the difference is greater than a preset threshold (such as 5% of the total workload), the position of the dividing line is adjusted; this process is repeated until the workload difference is less than the threshold or a preset number of iterations (such as 10 times) is reached. With three cleaning robots, the area was divided into three sub-areas with workloads of 105, 108, and 102 work units respectively, achieving a balanced distribution of workload.

[0170] After the sub-regions are divided, a cleaning trajectory is planned for each cleaning robot. An improved zigzag cleaning path is adopted to ensure 100% coverage. The cleaning trajectory is divided into high-density trajectory segments and low-density trajectory segments according to a preset dirt density threshold (e.g., density value 6).

[0171] For high-density trajectory segments, a first sliding surface is constructed. This sliding surface comprehensively considers velocity error and position error, with a velocity error weight of 0.7 and a position error weight of 0.3. This configuration allows the robot to maintain a low and stable speed in highly soiled areas, improving cleaning efficiency. For example, when the robot is cleaning in an area with a soil density of 8, the target speed is set to 0.1 m / s, the actual speed is 0.15 m / s, and the position deviation is 0.05 m. The calculated sliding surface state variable is 0.035, indicating that the control input needs to be reduced. For low-density trajectory segments, a second sliding surface is constructed. This sliding surface comprehensively considers the position error integral term and the position error itself, with the position error integral term weighted at 0.4 and the position error weighted at 0.6. This configuration allows the robot to pass quickly in low-soiled areas while maintaining trajectory accuracy. For example, when the robot is cleaning in an area with a soil density of 3, the cumulative position deviation is 0.1 m, and the current position deviation is 0.03 m. The calculated sliding surface state variable is 0.058, indicating that the control input needs to be adjusted appropriately. Based on the sliding surface state variables, equivalent control variables are calculated. These equivalent control variables are primarily used to compensate for the deterministic components of the robot's dynamics model, including parameters such as mass, inertia, and friction. For example, for a cleaning robot with a mass of 15 kg and a wheel spacing of 0.4 m, when the sliding surface state variable is 0.035, the calculated equivalent control variables are: left wheel motor torque 0.42 N·m and right wheel motor torque 0.38 N·m.

[0172] The adaptive switching gain is calculated based on dirt density, cleaning resistance, and sliding surface state variables. When the dirt density is high, the switching gain is increased to enhance disturbance rejection capability; when the cleaning resistance increases, the switching gain is also increased accordingly. For example, when the dirt density is 8, the cleaning resistance is 5 N, and the sliding surface state variable is 0.035, the calculated adaptive switching gain is 0.25.

[0173] The equivalent control quantity is combined with the sign function of the adaptive switching gain to generate a piecewise adaptive sliding mode control input. When the absolute value of the sliding surface state quantity is less than 0.01, a continuous approximation is used to replace the sign function to avoid control input jitter; when the absolute value of the state quantity is greater than 0.01, the standard sign function is used.

[0174] To avoid robot collisions, the percentage of cleaned area and distance to the target point for each cleaning robot are calculated in real time to determine obstacle avoidance priority. Robots with a higher percentage of cleaned area and closer distance to the target point have higher obstacle avoidance priority. For example, if robot A has a 75% cleaned area and is 0.8 meters from the target point, while robot B has a 60% cleaned area and is 1.5 meters from the target point, then robot A has a higher obstacle avoidance priority than robot B. When the distance between adjacent cleaning robots is less than a safe distance threshold (e.g., 0.5 meters), the segmented adaptive sliding mode control input is corrected based on the obstacle avoidance priority. The correction amount is calculated using the formula: Correction Amount = Basic Correction Coefficient × Priority Difference × Distance Gradient.

[0175] This invention achieves efficient collaboration in a multi-robot cleaning system through workload-balanced sub-region division and differentiated segmented sliding mode control strategies. Different control strategies are employed for areas with varying dirt densities, prioritizing cleaning quality in high-dirt areas and improving cleaning efficiency in low-dirt areas. Simultaneously, adaptive gain switching effectively suppresses external disturbances and parameter uncertainties during the cleaning process. An obstacle avoidance priority mechanism calculated based on the percentage of cleaned area and target distance effectively prevents robot collisions, enhancing the system's safety and reliability.

[0176] A second aspect of the present invention,

[0177] Provides an environment-aware adaptive cleaning path planning system for photovoltaic panels, including:

[0178] The first unit is used to acquire historical weather data and photovoltaic panel status monitoring data of the area where the photovoltaic panel is located, and calculate the cleanliness value of the photovoltaic panel; when the cleanliness value is lower than the cleaning threshold dynamically calculated based on power generation loss and cleaning cost, a cleaning trigger command is generated.

[0179] The second unit is used to receive the cleaning trigger command and execute the cleaning path optimization steps, including: setting the optimization objectives as maximizing cleaning efficiency, minimizing energy consumption, and optimizing time cost; dynamically adjusting the weights based on environmental impact and cleaning energy efficiency ratio; optimizing the population using a genetic algorithm based on dirt density encoding; mapping the optimized population to the initial distribution of cleaning efficiency pheromones, energy consumption pheromones, and time cost pheromones; obtaining the global path through iterative calculation using a multi-pheromone ant colony algorithm; and performing local optimization using an environmentally adaptive multi-neighborhood simulated annealing algorithm.

[0180] The third unit is used to execute multi-robot collaborative cleaning control steps based on the optimized cleaning path, including: dividing the area into sub-regions based on dirt density and workload balance; using segmented adaptive sliding mode variable structure control for trajectory tracking; and determining obstacle avoidance priority based on the percentage of cleaned area and target distance for multi-robot collaborative control.

[0181] A third aspect of the present invention,

[0182] An electronic device is provided, comprising:

[0183] processor;

[0184] Memory used to store processor-executable instructions;

[0185] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0186] Fourth aspect of the embodiments of the present invention,

[0187] A computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0188] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0189] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A photovoltaic panel adaptive cleaning path planning method based on environmental perception, characterized in that, include: Obtain historical weather data and photovoltaic panel status monitoring data for the area where the photovoltaic panels are located, and calculate the cleanliness value of the photovoltaic panels; When the cleanliness value is lower than the cleaning threshold dynamically calculated based on power generation loss and cleaning cost, a cleaning trigger command is generated; Upon receiving the cleaning trigger command, the cleaning path optimization steps are executed, including: setting the optimization objectives as maximizing cleaning efficiency, minimizing energy consumption, and optimizing time cost; dynamically adjusting the weights based on environmental impact and cleaning energy efficiency ratio; optimizing the population using a genetic algorithm based on dirt density encoding; mapping the optimized population to the initial distribution of cleaning efficiency pheromones, energy consumption pheromones, and time cost pheromones; obtaining the global path through iterative calculation using a multi-pheromone ant colony algorithm; and performing local optimization using an environmentally adaptive multi-neighborhood simulated annealing algorithm. Based on the optimized cleaning path, multi-robot collaborative cleaning control steps are executed, including: dividing the area into sub-regions based on dirt density and workload balance; using segmented adaptive sliding mode variable structure control for trajectory tracking; determining obstacle avoidance priority based on the percentage of cleaned area and target distance, and performing multi-robot collaborative control.

2. The method according to claim 1, characterized in that, Historical weather data and photovoltaic panel status monitoring data for the area where the photovoltaic panels are located are obtained, and the cleanliness value of the photovoltaic panels is calculated, including: Historical weather data is processed using a sliding time window method. A time decay factor is introduced into the data within each time window for weighted calculation to obtain the historical weather impact value. The historical weather impact values ​​and status monitoring data are multiplied by the corresponding weight coefficients and summed to obtain the photovoltaic panel cleanliness value. The initial values ​​of each weight coefficient are determined by fuzzy hierarchical analysis, and the gradient direction of the weights is dynamically adjusted based on the cleanliness value during the calculation process. Establish a historical cleaning record database to record the cleanliness value before each cleaning, cleaning cost, and actual power generation loss during the corresponding period. Combine seasonal and environmental factors to calculate the power generation loss caused by the reduction in unit cleanliness. A cleaning benefit evaluation function is established based on the unit cleaning cost and the power generation loss caused by the reduction in unit cleanliness. The corresponding cleaning benefit is calculated by traversing different cleanliness values, and the cleanliness value that maximizes the cleaning benefit is determined as the cleaning threshold for the current period. The cleaning benefit evaluation function also considers the cost of cleaning resource scheduling and the time window constraint of cleaning operation.

3. The method according to claim 1, characterized in that, The weights are dynamically adjusted based on environmental impact and cleaning energy efficiency ratio, including: An environmental impact function is established to characterize the effects of temperature, humidity, wind speed, light intensity, and precipitation probability on the cleaning effect. The temperature parameter is calculated using an exponential function to determine the degree of influence of temperature deviation from the optimal cleaning temperature, and the humidity parameter is calculated using an S-shaped response curve to determine the degree of influence of humidity exceeding a critical value. A coupling coefficient between the temperature and humidity parameters is constructed based on the environmental impact function. The grid cells are divided based on the density of dirt distribution on the photovoltaic panel surface, and initial weights are set for cleaning efficiency, energy consumption, and time cost. An environmental adjustment coefficient is calculated based on the environmental impact function and the coupling coefficient, and the environmental adjustment coefficient is positively correlated with the environmental sensitivity of the grid cell; an energy efficiency ratio adjustment coefficient is calculated based on the cleaning energy efficiency ratio of the grid cell, and the energy efficiency ratio adjustment coefficient is a power function relationship between the current cleaning energy efficiency ratio and the maximum cleaning energy efficiency ratio; The product of the environmental adjustment coefficient and the energy efficiency ratio adjustment coefficient is used as the weight adjustment amount, so that the cleaning efficiency weight is negatively correlated with the weight adjustment amount, and the energy consumption weight and the time cost weight are positively correlated with the weight adjustment amount, and the initial weights are dynamically updated.

4. The method according to claim 1, characterized in that, A genetic algorithm based on dirt density encoding is used to optimize the population, mapping the optimized population to the initial distribution of cleaning efficiency pheromones, energy consumption pheromones, and time cost pheromones. A global path is obtained through iterative calculation using a multi-pheromone ant colony algorithm, and local optimization is performed using an environment-adaptive multi-neighborhood simulated annealing algorithm, including: The initial population is initialized and constructed using a genetic algorithm, which employs an encoding strategy based on dirt density weights and an environment-aware adaptive crossover and mutation mechanism. The optimization results of the genetic algorithm are used as the initial pheromone distribution of the ant colony algorithm. A multi-pheromone ant colony algorithm based on grid energy efficiency gradient is used to generate the globally optimal cleaning path. The multi-pheromone includes cleaning efficiency pheromone, energy consumption pheromone and time cost pheromone. The globally optimal cleaning path is locally optimized using an environment-adaptive multi-neighborhood simulated annealing algorithm. The multi-neighborhood includes reverse neighborhood, exchange neighborhood, insertion neighborhood, and region reconstruction neighborhood. Local path optimization is performed through an environment-aware temperature adjustment mechanism and a dynamic acceptance criterion strategy.

5. The method according to claim 4, characterized in that, The genetic algorithm employs an encoding strategy based on dirt density weights and an environment-aware adaptive crossover and mutation mechanism, including: The characteristics of various types of dirt on the surface of the photovoltaic panel are obtained, the surface of the photovoltaic panel is divided into grid cells, and the dirt density of each grid cell is calculated. The chromosome of the genetic algorithm is constructed as a multi-level coding structure of spatial sequence coding, resource allocation coding and density correlation coding, where spatial sequence coding represents the cleaning path, resource allocation coding represents the cleaning parameters, and density correlation coding represents the dirt handling strategy. A density weight mapping function is constructed based on the dirt density. The density weight mapping function simultaneously considers the dirt density of the target grid cell and the dirt density distribution of its adjacent grid cells. The population is initialized according to the density weight mapping function. Environmental parameters are acquired and environmental impact factors are calculated. Based on the environmental impact factors and the density weight mapping function, the crossover probability and mutation probability of the genetic algorithm are dynamically adjusted. The cleaning path sequence is optimized through crossover operation, and the resource allocation strategy is adjusted through mutation operation. The spatial sequence encoding, resource allocation encoding, and density correlation encoding of the best individual in the optimized population are mapped to the initial pheromone distribution matrix of the ant colony algorithm.

6. The method according to claim 5, characterized in that, The optimization results of the genetic algorithm are used as the initial pheromone distribution for the ant colony algorithm. A multi-pheromone ant colony algorithm based on grid energy efficiency gradient is used to generate the globally optimal cleaning path. The multi-pheromone includes cleaning efficiency pheromone, energy consumption pheromone, and time cost pheromone. The optimization result of the genetic algorithm is mapped to a multidimensional pheromone initial distribution, which includes cleaning efficiency pheromone, energy consumption pheromone and time cost pheromone. Each pheromone component is obtained by weighted combination of the sequence information corresponding to the optimal individual and the grid dirt density. The grid energy efficiency ratio is calculated based on the ratio of cleaning revenue to energy consumption of the grid cells, and the grid energy efficiency gradient vector is constructed based on the partial derivatives of the grid energy efficiency ratio in the horizontal and vertical directions. The heuristic information of each pheromone component is weighted and multiplied to obtain the comprehensive heuristic information. Based on the comprehensive heuristic information, the initial distribution of multidimensional pheromones, and the grid energy efficiency gradient vector, the state transition probability matrix is ​​calculated. The ant's movement path is determined based on the state transition probability matrix. Local pheromone updates are performed on the grid cells along the movement path, taking into account the influence of the pheromone evaporation coefficient and the grid energy efficiency ratio. The global optimal path is then enhanced with pheromone based on the evaluation index of the movement path. Calculate the variance of each pheromone component, perform multidimensional convergence determination based on the relative rate of change of the variance, and output the globally optimal cleaning path when the multidimensional convergence determination result is less than a preset convergence threshold.

7. The method according to claim 1, characterized in that, The multi-robot collaborative cleaning control steps include: The dirt density distribution and area of ​​the cleaning area are obtained, the workload of the sub-area is calculated, and the cleaning area is divided into multiple sub-areas and assigned to multiple cleaning robots based on the balance of the workload of the sub-areas. The cleaning trajectory in each sub-region is divided into high-density trajectory segments and low-density trajectory segments according to the dirt density threshold. A first sliding surface is constructed by weighted combination of velocity error and position error for the high-density trajectory segments, and a second sliding surface is constructed by weighted combination of position error integral term and position error for the low-density trajectory segments. The equivalent control quantity is calculated based on the state quantities of the first and second sliding surfaces, and the equivalent control quantity is used to compensate for the deterministic dynamic characteristics during the cleaning process; the adaptive switching gain is calculated based on the dirt density, cleaning resistance and the state quantities of the sliding surfaces, and the adaptive switching gain is used to suppress uncertain disturbances during the cleaning process; the equivalent control quantity and the sign function of the adaptive switching gain are combined to generate a piecewise adaptive sliding mode control input. Calculate the percentage of cleaned area and distance from the target point for each cleaning robot to determine the obstacle avoidance priority; when the distance between adjacent cleaning robots is less than the safe distance threshold, the segmented adaptive sliding mode control input is corrected according to the obstacle avoidance priority, and the correction amount is proportional to the difference in obstacle avoidance priority between adjacent robots and the distance gradient.

8. An environmentally-aware adaptive cleaning path planning system for photovoltaic panels, used to implement the method described in any one of claims 1-7, characterized in that, include: The first unit is used to acquire historical weather data and photovoltaic panel status monitoring data for the area where the photovoltaic panels are located, and to calculate the cleanliness value of the photovoltaic panels. When the cleanliness value is lower than the cleaning threshold dynamically calculated based on power generation loss and cleaning cost, a cleaning trigger command is generated; The second unit is used to receive the cleaning trigger command and execute the cleaning path optimization steps, including: setting the optimization objectives as maximizing cleaning efficiency, minimizing energy consumption, and optimizing time cost; dynamically adjusting the weights based on environmental impact and cleaning energy efficiency ratio; optimizing the population using a genetic algorithm based on dirt density encoding; mapping the optimized population to the initial distribution of cleaning efficiency pheromones, energy consumption pheromones, and time cost pheromones; obtaining the global path through iterative calculation using a multi-pheromone ant colony algorithm; and performing local optimization using an environmentally adaptive multi-neighborhood simulated annealing algorithm. The third unit is used to execute multi-robot collaborative cleaning control steps based on the optimized cleaning path, including: dividing the area into sub-regions based on dirt density and workload balance; using segmented adaptive sliding mode variable structure control for trajectory tracking; and determining obstacle avoidance priority based on the percentage of cleaned area and target distance for multi-robot collaborative control.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Images