Photovoltaic cleaning path and water quantity scheduling method and system based on turbidity feedback

Abstract

The application discloses a photovoltaic cleaning path and water quantity scheduling method and system based on turbidity feedback, and belongs to the technical field of photovoltaic power station operation and maintenance. The method generates a panel-level pollution list by collecting photovoltaic panel surface images and panel optical signals, determines panel-level target water quantity by using a preset mapping table, generates a batch water consumption budget table and a panel-level water distribution table in combination with water tank capacity, and plans an operation path with the start and end points as the water supplement positions. The turbidity and water quantity of the backflow water are collected in real time during the cleaning process, and the cleaning standard is determined according to the comparison with the turbidity threshold value. The residual water quantity of the photovoltaic panel reaching the standard in advance is recovered, and the residual water of the batch is updated. The water influence degree and panel-level water demand of the photovoltaic panel not reaching the standard are calculated, and the re-cleaning object is selected and inserted into the re-cleaning path under the constraint of the batch water allowance. The application realizes fine panel-level water control and re-cleaning optimization scheduling, improves the unit water quantity utilization efficiency and the cleaning standard rate, and reduces the labor participation and operation and maintenance cost.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power plant operation and maintenance technology, specifically to a photovoltaic cleaning path and water volume scheduling method and system based on turbidity feedback. Background Technology

[0002] Photovoltaic power plants are typically located in open areas such as deserts, tidal flats, or mountains, with a large number of modules arranged in rows. Affected by wind and sand deposition, rainwater redeposition, bird activity, and construction dust, the surface fouling of the modules exhibits significant non-uniformity in space and time, leading to a decrease in power generation and an increased risk of localized hot spots.

[0003] The cleaning strategies commonly used in current operation and maintenance often use fixed cycles, weather experience, or power generation attenuation thresholds as triggering conditions; cleaning equipment (including cleaning robots) usually perform operations in rows and columns, and spraying / scrubbing parameters are given by preset duration, pressure, or flow rate levels, and water budgets are mostly statically allocated based on water tank capacity or experience quotas.

[0004] Due to the lack of online closed-loop indicators that can quantify the cleaning effect during the operation, the need for re-sweeping is often determined by manual spot checks, image comparison, or power recovery after the operation. When the water tank capacity is limited and the water replenishment point is constrained by the site conditions, static quotas are prone to conservative restrictions on high-efficiency objects and repeated consumption of low-efficiency objects, making it difficult to simultaneously consider the compliance rate, walking distance, and unit water volume benefits within a single water replenishment cycle. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a photovoltaic cleaning path and water volume scheduling method based on turbidity feedback, comprising the following steps: Step 1: Collect information on the surface contamination of photovoltaic panels and generate a panel-level contamination list; generate a batch water usage budget table and a panel-level water usage allocation table based on the panel-level contamination list, and generate a batch operation path based on the batch water usage budget table and the panel-level water usage allocation table; during the cleaning operation of each photovoltaic panel along the batch operation path, collect the turbidity of the return water of each photovoltaic panel in real time, and record the actual water consumption and operation time of each photovoltaic panel; Step 2: Compare the turbidity of the return water from each photovoltaic panel with a preset return water turbidity threshold. When the turbidity of the return water from a photovoltaic panel is less than or equal to the threshold, stop cleaning that panel and determine that it has met the cleaning standard. Add the target remaining water volume of that photovoltaic panel to the reserved water volume in the batch water budget table and update the batch remaining water volume in the batch water budget table. When the turbidity of the return water is still not less than the threshold after the corresponding spraying time has been reached, stop cleaning that photovoltaic panel, determine that it has not met the cleaning standard, and add the corresponding photovoltaic panel to the batch re-cleaning candidate set. The target remaining water volume is the difference between the target water volume for each panel and its actual water volume. When the difference is less than zero, the target remaining water volume is zero. Step 3: Obtain the turbidity difference between the turbidity of the return water at the time of stopping cleaning for each non-compliant photovoltaic panel and the turbidity threshold value of the return water. Obtain the turbidity reduction value between the maximum value of the return water turbidity during the cleaning process and the turbidity of the return water at the time of stopping cleaning for each non-compliant photovoltaic panel. Based on the turbidity difference and the turbidity reduction value, and using the larger of the turbidity difference and the preset minimum difference value as the denominator, calculate the ratio of the turbidity reduction value to the denominator as the water use impact degree of the corresponding photovoltaic panel. Based on the product of the panel-level target water consumption and the water use impact degree of each non-compliant photovoltaic panel, obtain the panel-level water demand of each non-compliant photovoltaic panel. Step 4: Calculate the water margin for this batch based on the updated remaining water volume in the batch water budget table, and compare the sum of the water demand for each non-compliant photovoltaic panel with the remaining water volume of the batch: when the sum is less than the remaining water volume of the batch, the water demand for each non-compliant photovoltaic panel is directly used as the basis for generating the water allocation table for non-compliant photovoltaic panels; when the sum is not less than the remaining water volume of the batch, the corresponding non-compliant photovoltaic panels are removed in order of increasing water impact until the sum of the water demand for the remaining non-compliant photovoltaic panels is less than the remaining water volume of the batch, and then the water allocation table for the non-compliant photovoltaic panels is generated based on the remaining non-compliant photovoltaic panels. Step 5: Based on the panel-level water allocation table of the substandard photovoltaic panels, insert a re-cleaning path before the end of the batch operation path; after the re-cleaning path is completed, end the current batch, replenish water, and clean the next batch until the operation task is completed.

[0006] Furthermore, the process involves collecting photovoltaic panel surface contamination information to obtain a panel-level contamination list, generating a batch water usage budget table and a panel-level water usage allocation table based on the panel-level contamination list, and obtaining a batch operation path based on the batch water usage budget table and the panel-level water usage allocation table, including: The system collects images of the photovoltaic panel surface, optical signals from the panel surface, and ambient light intensity, extracts dirt features, generates dirt severity values ​​and stain type labels, and forms a panel-level dirt list. Based on the panel-level dirt list, the system determines the target water consumption for each photovoltaic panel from a preset mapping table. The system uses the water tank capacity of the cleaning robot as the upper limit of water consumption per batch to divide the work tasks into batches, generating a batch water consumption budget table and a panel-level water consumption allocation table. Based on the panel-level water consumption allocation table, the water replenishment location is set as the start and end points of the batch operation path, generating the batch operation path.

[0007] Furthermore, the process involves acquiring images of the board surface, optical signals from the board surface, and ambient light intensity; extracting soiling features; generating soiling severity values ​​and stain type labels; and forming a board-level soiling list, including: The photovoltaic panel data acquisition module collects image data and optical signals of the photovoltaic panel surface. The stain type is identified based on the set stain type identification features and the image data. The optical signal is the reflected light intensity obtained when light of a preset intensity shines on the photovoltaic panel. Based on a pre-calibrated reflection intensity-stain degree mapping relationship, the stain degree value is obtained according to the reflection intensity. The stain degree value and the stain type are used to form a panel-level stain list.

[0008] Furthermore, based on the panel-level contamination list, the target water consumption for each photovoltaic panel is determined from a preset mapping table. The work tasks are divided into batches using the water tank capacity as the upper limit for water consumption in a single batch, generating a batch water consumption budget table and a panel-level water consumption allocation table, including: Read the nozzle opening corresponding to the stain type and the spray duration corresponding to the degree of staining from the preset mapping table, and determine the target water consumption of the plate level based on the spray duration and the nozzle opening; The cleaning robot's water tank capacity is used as the upper limit for water consumption per batch to divide the work tasks into batches, forming a batch water consumption budget table that includes allocated water consumption and reserved water consumption. The reserved water consumption is used for re-cleaning and abnormal consumption in this batch. Based on the allocated water consumption and the target water consumption for each panel, the photovoltaic panels that can be cleaned in this batch are determined, and a panel-level water consumption allocation table is formed based on the information of the cleanable photovoltaic panels, the spraying duration, the nozzle opening, and the target water consumption for each panel. The sum of the allocated water consumption and the reserved water consumption is equal to the water tank capacity. The panel-level water consumption allocation table corresponds one-to-one with the spraying duration and nozzle opening parameters and the target water consumption for each panel.

[0009] Furthermore, the step of setting the water replenishment location as the start and end point of the batch operation path based on the plate-level water allocation table, and generating the batch operation path, includes: Obtain the location of each photovoltaic panel in the panel-level water allocation table to obtain the arrangement order of the photovoltaic panel locations. Add water replenishment locations at the beginning and end of the arrangement order of the photovoltaic panel locations to obtain the batch operation node order. Generate the batch operation path according to the batch operation node order.

[0010] Furthermore, the step of calculating the water leeway for this batch based on the updated remaining water volume in the batch water budget table includes: Calculate the total water demand of each non-compliant photovoltaic panel; if the total is less than the remaining water demand of the batch, the water demand of this batch is determined to be sufficient; otherwise, the water demand of this batch is determined to be insufficient.

[0011] Furthermore, the panel-level water allocation table based on substandard photovoltaic panels inserts a re-cleaning path before the end of the batch operation path, including: The sequence of operation nodes for the substandard photovoltaic panels is determined based on the panel-level water allocation table, a re-cleaning path is generated, and the re-cleaning path is inserted before the end point of the batch operation path.

[0012] Furthermore, the step of ending the current batch after completing the re-cleaning path also includes: The non-compliant photovoltaic panels that were rejected in each batch were compiled into a legacy list. The photovoltaic panels in the legacy list were either manually processed in a unified manner or cleaned separately.

[0013] The photovoltaic cleaning robot cleaning system, which applies the photovoltaic cleaning path and water volume scheduling method based on turbidity feedback, includes: a data acquisition unit, a soiling assessment unit, a mapping table management and water usage budget unit, a batch path planning unit, an execution control and turbidity acquisition unit, a compliance assessment unit, a water usage margin calculation and re-cleaning scheduling unit, a scheduling and result output unit, and a data processing module. The data acquisition unit, pollution assessment unit, mapping table management and water budgeting unit, batch path planning unit, execution control and turbidity acquisition unit, compliance assessment unit, water margin calculation and re-cleaning scheduling unit, and scheduling and result output unit are respectively connected to the data processing module. The data acquisition unit is used to acquire the board surface image, board surface optical signal and ambient light intensity; The soiling assessment unit is used to extract soiling features and generate soiling degree values ​​and stain type labels to form a board-level soiling list. The mapping table management and water use budget unit is used to determine the target water consumption at the board level from the preset mapping table, and generate a batch water use budget table and a board level water use allocation table, which include the allocated water consumption and the reserved water consumption, with the water tank capacity as the upper limit of water consumption in a single batch. The batch path planning unit is used to set the water replenishment location as the start and end point of the batch operation path and generate the batch operation path. The execution control and turbidity acquisition unit is used to perform spraying and brushing along the batch operation path, control the cleaning according to the plate water allocation table, collect the turbidity of the return water in real time, and record the actual water consumption and operation time. The compliance assessment unit is used to conduct compliance assessment based on the turbidity of the return water and the turbidity threshold of the return water, and output a batch re-cleaning candidate set. The water margin calculation and re-cleaning scheduling unit is used to calculate the water margin of this batch based on the batch re-cleaning candidate set, the batch water budget table and the cumulative water consumption of this batch, generate the panel-level water allocation table for non-compliant photovoltaic panels in combination with the water impact degree, and insert the re-cleaning path before the end of the batch operation path. The scheduling and result output unit is used to trigger water replenishment after the current batch ends and continue cleaning the remaining batches, while outputting the batch water usage log, the result of the return water turbidity meeting the standard, and the re-cleaning record.

[0014] The beneficial effects of this invention are: by using a panel-level contamination list and a preset mapping table, the type and degree of contamination are directly associated with the nozzle opening, spraying duration and panel-level target water consumption, thereby achieving personalized water use planning for each photovoltaic panel.

[0015] The system uses real-time turbidity of the return water and set thresholds to determine whether the cleaning meets the standards, and dynamically adjusts the remaining water consumption of each batch by recovering the target remaining water consumption, thereby improving the efficiency of water resource use.

[0016] By comparing the turbidity difference and the turbidity reduction value, the water use impact is defined. The benefit priority of water reuse is determined from the perspective of the turbidity improvement trend, and the limited remaining water use is preferentially allocated to photovoltaic panels with higher water use impact.

[0017] By comparing the overall water demand of substandard photovoltaic panels with the remaining water volume of the batch, the system automatically determines the water leeway status and removes panels with a lower impact on water usage when water leeway is insufficient, thus achieving the optimal trade-off between water usage and cleaning quality.

[0018] In batch path planning, the water replenishment location is set as the start and end point, and a re-cleaning path is inserted before the end point to reduce repeated travel and idle distances, thereby improving the robot's operating efficiency. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating the photovoltaic cleaning path and water volume scheduling method based on turbidity feedback. Detailed Implementation

[0020] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the following description.

[0021] The features and performance of the present invention will be further described in detail below with reference to embodiments.

[0022] Example 1 like Figure 1 As shown, the photovoltaic cleaning path and water volume scheduling method based on turbidity feedback includes the following steps: Step 1: Collect information on the surface contamination of photovoltaic panels and generate a panel-level contamination list; generate a batch water usage budget table and a panel-level water usage allocation table based on the panel-level contamination list, and generate a batch operation path based on the batch water usage budget table and the panel-level water usage allocation table; during the cleaning operation of each photovoltaic panel along the batch operation path, collect the turbidity of the return water of each photovoltaic panel in real time, and record the actual water consumption and operation time of each photovoltaic panel; Step 2: Compare the turbidity of the return water from each photovoltaic panel with a preset return water turbidity threshold. When the turbidity of the return water from a photovoltaic panel is less than or equal to the threshold, stop cleaning that panel and determine that it has met the cleaning standard. Add the target remaining water volume of that photovoltaic panel to the reserved water volume in the batch water budget table and update the batch remaining water volume in the batch water budget table. When the turbidity of the return water is still not less than the threshold after the corresponding spraying time has been reached, stop cleaning that photovoltaic panel, determine that it has not met the cleaning standard, and add the corresponding photovoltaic panel to the batch re-cleaning candidate set. The target remaining water volume is the difference between the target water volume for each panel and its actual water volume. When the difference is less than zero, the target remaining water volume is zero. Step 3: Obtain the turbidity difference between the turbidity of the return water at the time of stopping cleaning for each non-compliant photovoltaic panel and the turbidity threshold value of the return water. Obtain the turbidity reduction value between the maximum value of the return water turbidity during the cleaning process and the turbidity of the return water at the time of stopping cleaning for each non-compliant photovoltaic panel. Based on the turbidity difference and the turbidity reduction value, and using the larger of the turbidity difference and the preset minimum difference value as the denominator, calculate the ratio of the turbidity reduction value to the denominator as the water use impact degree of the corresponding photovoltaic panel. Based on the product of the panel-level target water consumption and the water use impact degree of each non-compliant photovoltaic panel, obtain the panel-level water demand of each non-compliant photovoltaic panel. Step 4: Calculate the water margin for this batch based on the updated remaining water volume in the batch water budget table, and compare the sum of the water demand for each non-compliant photovoltaic panel with the remaining water volume of the batch: when the sum is less than the remaining water volume of the batch, the water demand for each non-compliant photovoltaic panel is directly used as the basis for generating the water allocation table for non-compliant photovoltaic panels; when the sum is not less than the remaining water volume of the batch, the corresponding non-compliant photovoltaic panels are removed in order of increasing water impact until the sum of the water demand for the remaining non-compliant photovoltaic panels is less than the remaining water volume of the batch, and then the water allocation table for the non-compliant photovoltaic panels is generated based on the remaining non-compliant photovoltaic panels. Step 5: Based on the panel-level water allocation table of the substandard photovoltaic panels, insert a re-cleaning path before the end of the batch operation path; after the re-cleaning path is completed, end the current batch, replenish water, and clean the next batch until the operation task is completed.

[0023] First, the cleaning robot or its associated data acquisition equipment acquires images of the photovoltaic panel surface, optical signals of the panel surface, and ambient light intensity.

[0024] The photovoltaic panel data acquisition module includes a camera device for acquiring image data of the photovoltaic panel surface. This image data can at least reflect the dust distribution, stain morphology, bird droppings, and other dirt on the photovoltaic panel surface.

[0025] The optical signal of the photovoltaic panel is acquired through a combination of a light source and a photosensitive device: the system applies a preset intensity of light (which can be a light source with fixed brightness and a fixed spectral combination) to the surface of the photovoltaic panel; the photosensitive device collects the light intensity signal reflected back from the surface of the photovoltaic panel. This reflected light intensity can be regarded as the optical signal of the panel surface, used to characterize the reflective properties of the panel surface, and the degree of surface contamination is estimated by combining it with the calibrated reflection intensity-contamination degree mapping relationship.

[0026] An ambient light intensity sensor is used to measure the current natural light intensity and to perform background subtraction, normalization, or drift correction on the optical signal of the panel surface to reduce the impact of natural light fluctuations on the contamination assessment.

[0027] The data processing module processes the aforementioned data, including: identifying stained areas in the panel image and analyzing the area, distribution, shape, and color of the stains; classifying the stains into preset stain types, such as dust, mud, and bird droppings, based on their color and morphological characteristics, and assigning one or more stain type labels to each photovoltaic panel; and estimating the overall degree of shading on the panel by combining the panel's optical signals and ambient light intensity. For example, when the reflected light intensity is significantly reduced and the corresponding area has dense dust, the degree of contamination can be determined to be high.

[0028] To facilitate subsequent processing, this embodiment generates a soiling level value for each photovoltaic panel, typically categorized into multiple levels such as light, moderate, and heavy. In practice, factors affecting light transmission, such as the shading ratio, the reduction in reflected light intensity, and the area covered by stains, can be combined using pre-defined rules to obtain a level index representing the overall soiling level. Simultaneously, the corresponding stain type label is combined with the soiling level value to form a soiling record for each photovoltaic panel.

[0029] By identifying and assessing all photovoltaic panels in the entire area to be cleaned, a panel-level dirt list is obtained. This list should include at least: the photovoltaic panel number or location identifier, the corresponding stain type label, and the corresponding dirt level value.

[0030] The system pre-establishes a mapping table to map stain type and degree of soiling to nozzle opening and spray duration.

[0031] For different types of stains, such as dust, the spraying time requirement may be low, while bird droppings may require both nozzle opening and spraying time to be high. For different levels of soiling severity, the higher the soiling severity, the greater the requirements for spray duration and / or nozzle opening. The mapping table, configured through experimentation or experience, can provide corresponding spray parameters for each combination of stain type and soiling severity value.

[0032] For each photovoltaic panel: First, the system looks up the nozzle opening corresponding to its stain type in the mapping table; then, it looks up the spraying duration corresponding to its stain severity value; based on the water consumption per unit time of the nozzle at that opening and the corresponding spraying duration, the system determines the panel-level target water consumption for that photovoltaic panel. Specifically, this includes: first, determining the amount of water sprayed per unit time (e.g., per second or per minute) by the nozzle at the current opening setting; then, based on the spraying duration of the photovoltaic panel, calculating the total amount of water sprayed within that duration, which is defined as the panel-level target water consumption for that photovoltaic panel.

[0033] Therefore, a target water consumption can be determined for each photovoltaic panel in the panel-level pollution inventory.

[0034] The water tank of the cleaning robot has a fixed capacity, which is used as the upper limit for water consumption during a single batch of operations.

[0035] During the planning phase, the system divides batches according to the following logic: an allocated water volume and a reserved water volume are set for the current batch; the allocated water volume is used to support the initial cleaning of each photovoltaic panel in the current batch; the reserved water volume is used in advance for possible re-cleaning and abnormal water use in this batch (such as local stubborn stains, additional water use caused by nozzle malfunction, etc.); the sum of the allocated water volume and the reserved water volume shall not exceed the water tank capacity, and is usually equal to the water tank capacity.

[0036] In practical implementation, it can be planned as follows: Based on the total target water consumption of all photovoltaic panels in the overall task, and in conjunction with the water tank capacity, determine how many batches are needed. In each batch, select and allocate photovoltaic panels according to the target water consumption at the panel level, so that the total allocated water consumption does not exceed the upper limit of the allocated water consumption available in that batch, and reserve a certain proportion or a certain fixed amount of water as reserve water consumption.

[0037] This results in a batch water usage budget, which includes at least: the allocated water usage for each batch, the reserved water usage for each batch, and the set of photovoltaic panels allocated to each batch.

[0038] For each batch of photovoltaic panels, the system generates a panel-level water allocation table based on the panel-level target water consumption, spray duration, and nozzle opening. This table records for each photovoltaic panel: panel location or number, target spray duration, nozzle opening, and target water consumption per panel. The panel-level water allocation table serves as the source of control parameters for subsequent cleaning processes.

[0039] The system generates batch operation paths based on the location information of each photovoltaic panel in the panel-level water allocation table: The water replenishment location is set as the start and end point of the batch operation path. That is, the robot starts from the water replenishment point, completes the cleaning of all photovoltaic panels in the batch, and returns to the water replenishment point. The arrangement order of the photovoltaic panel positions can be arranged by distance priority, row and column order or other optimization algorithms to form a path that continuously visits all target photovoltaic panels. The water replenishment location is inserted at the beginning and end of the path to obtain the order of batch operation nodes, and then the batch operation path is generated accordingly.

[0040] During the execution phase, the robot sequentially reaches each photovoltaic panel according to the batch operation path and controls the spraying and washing according to the panel-level water allocation table: When the robot reaches a photovoltaic panel, it sets the spraying parameters according to the nozzle opening of the photovoltaic panel and starts the spraying and scrubbing action. During the spraying process, a turbidity sensor is used to collect the turbidity of the water flowing back from the surface of the photovoltaic panel to the recycling system in real time, i.e., the turbidity of the return water; the system also records the amount of water used by the photovoltaic panel and the cumulative operating time.

[0041] For each photovoltaic panel, the actual spraying time of the robot may be equal to the target spraying time for the panel, or it may stop early after the target conditions are met.

[0042] The system presets a return water turbidity threshold to determine whether the cleanliness of the photovoltaic panel surface meets the standard: during the cleaning process, the system compares the current return water turbidity with the return water turbidity threshold; when a photovoltaic panel has not yet reached its target spraying time, if its return water turbidity is less than or equal to the return water turbidity threshold, it indicates that the stains on the photovoltaic panel have been basically removed, and the marginal benefit of continuing to spray is low.

[0043] The system then performs the following operations: immediately stops spraying and scrubbing the photovoltaic panel; determines that the photovoltaic panel has met the cleaning standards; and calculates the target remaining water consumption for the photovoltaic panel. The target remaining water consumption can be understood as the non-negative difference between the originally planned target water consumption of the photovoltaic panel and the actual water consumption, that is, the portion of water that should have been available but was not actually used; if the actual water consumption exceeds the target water consumption of the panel, the target remaining water consumption is zero.

[0044] The target remaining water consumption of the photovoltaic panel is added to the reserved water consumption of the current batch to obtain the updated total reserved water consumption.

[0045] The remaining water consumption of the current batch is updated based on the water tank capacity, the cumulative actual water consumption of the current batch, and the updated total reserved water consumption. The remaining water consumption of the batch is the water tank capacity minus the cumulative actual water consumption of the current batch minus the updated total reserved water consumption. When the calculation result is less than zero, the remaining water consumption of the batch is zero.

[0046] The remaining water volume in a batch is used to characterize the amount of available water that can still be freely allocated within the current batch without occupying the reserved water volume, and serves as a constraint for subsequent water adequacy determination and water allocation for re-cleaning.

[0047] If the turbidity of the return water is still not less than or equal to the turbidity threshold of the return water after a certain photovoltaic panel has been sprayed for the target spraying time, it indicates that the water consumption according to the original plan is not enough to completely remove the stains.

[0048] At this time, the system performs the following operations: stops spraying and brushing the photovoltaic panel; determines that the cleaning of the photovoltaic panel has not met the standards; and adds the photovoltaic panel to the current batch of re-cleaning candidate set to provide a basis for subsequent re-cleaning planning.

[0049] For each non-compliant photovoltaic panel in the candidate set for re-cleaning, the system calculates its water use impact and panel-level water demand based on the change curve of return water turbidity over time.

[0050] For a non-compliant photovoltaic panel: the system records the turbidity of the return water when cleaning stops; calculates the difference between the turbidity of the return water and the turbidity threshold of the return water, and records it as the turbidity difference, which is used to reflect how much distance there is from meeting the standard. During the cleaning process of the photovoltaic panels, the historical changes in the turbidity of the return water are recorded, and the maximum value of the return water turbidity is found. The difference between this maximum value and the turbidity of the return water when cleaning stops is calculated and recorded as the turbidity reduction value, which is used to reflect the extent to which the turbidity has been reduced during this cleaning process.

[0051] This invention defines the degree of water use impact by measuring the ratio between the turbidity difference and the turbidity reduction value. If the turbidity reduction is significant during a period of water usage, while the turbidity difference when the current standard is not yet met is relatively small, it indicates that increasing water usage during this cleaning process has a significant effect on improving turbidity; the impact of water usage on this type of photovoltaic panel is relatively high. Conversely, if the turbidity reduction value is small while the turbidity difference value is still large, it indicates that even if a certain amount of water has been used, the improvement in turbidity is still limited; such photovoltaic panels will have low benefits from further increasing water use and low impact from water use.

[0052] In practical implementation, the system first calculates the turbidity difference and turbidity reduction value for each non-compliant photovoltaic panel, and uses the larger of the turbidity difference and the preset minimum difference as the denominator. The ratio of the turbidity reduction value to the denominator is then used as the numerical water use impact to avoid numerical anomalies caused by excessively small turbidity differences.

[0053] A greater impact of water use indicates a higher likelihood of continued reduction in turbidity and higher marginal benefits under the same additional water use conditions; a smaller impact of water use indicates lower benefits from increased water use.

[0054] After obtaining the water use impact level, the system further combines the target water use at the panel level to calculate the panel-level water demand for each non-compliant photovoltaic panel: For photovoltaic panels that are highly affected by water use, it is indicated that further increasing water use may significantly improve turbidity, and the water demand at the panel level should be appropriately increased. For photovoltaic panels with low water usage impact, it means that even if water usage is increased, it is difficult to effectively reduce turbidity. Therefore, the water usage requirement at the panel level can be relatively reduced or no additional water usage should be allocated.

[0055] In this embodiment, the target water consumption at the slab level and the water consumption impact can be combined according to a multiplicative relationship to obtain the water consumption demand at the slab level: the higher the water consumption impact, the greater the demand obtained based on the same target water consumption at the slab level.

[0056] The system summarizes the water demand at the panel level for all non-compliant photovoltaic panels in the current batch: Calculate the total water demand for each non-compliant photovoltaic panel; compare this total with the remaining water demand for the current batch: If the total is less than the remaining water consumption of the batch, then the water consumption margin for this batch is considered sufficient. If the total is not less than the remaining water consumption of the batch, then the water consumption margin of this batch is considered insufficient.

[0057] When the water slack is sufficient, the remaining water volume of the current batch is enough to meet the expected water demand of all non-compliant photovoltaic panels. At this time, the system directly uses the panel-level water demand of each non-compliant photovoltaic panel as the basis for generating the panel-level water allocation table, and arranges the corresponding re-cleaning water volume for each non-compliant photovoltaic panel. When the water slack is insufficient, the remaining water volume of the current batch is not enough to fully meet the water demand of all non-compliant photovoltaic panels. In this case, the system will select the appropriate option as follows: Sort all non-compliant photovoltaic panels in ascending order of water consumption impact; starting with the photovoltaic panel with the least water consumption impact, remove them from the re-cleaning plan in sequence; after each photovoltaic panel is removed, recalculate the total water consumption of the remaining non-compliant photovoltaic panels; continue until the total water consumption of the remaining non-compliant photovoltaic panels is less than the remaining water consumption of the batch.

[0058] Through the above steps, the system can prioritize the allocation of limited water resources to photovoltaic panels with higher water usage impact under limited water conditions, thereby improving the overall cleanliness of water usage per unit.

[0059] For the non-compliant photovoltaic panels that are ultimately retained in the re-cleaning plan, the system generates a panel-level water allocation table based on their panel-level water demand, recording the following for each photovoltaic panel: The planned water usage for subsequent cleaning; the corresponding spraying parameters (nozzle opening and spraying duration adjustment scheme); and information such as the location of the photovoltaic panels.

[0060] After obtaining the water allocation table for the substandard photovoltaic panels, the system plans a re-cleaning path that includes these panels based on their location information.

[0061] The sequence of operation nodes in the re-cleaning path can be optimized based on the relative positions of the non-compliant photovoltaic panels, for example, by adopting the shortest path priority principle; After generating the re-cleaning path, insert the path before the end point of the original batch cleaning path, so that after the robot completes the first round of cleaning, it will first perform re-cleaning and then return to the water replenishment point.

[0062] After completing all cleaning tasks along the batch operation path following the inserted cleaning path, the robot returns to the water replenishment position and ends the current batch operation.

[0063] At this point, the system performs the following operations: Update the batch water usage log based on the compliance and water usage of each photovoltaic panel within the batch; record the non-compliant photovoltaic panels that were removed from the batch due to insufficient water supply as a legacy list for subsequent manual handling or separate cleaning when conditions permit; replenish the robot's water tank at the water replenishment location, preparing for the next batch. These steps are repeated multiple times in a cycle until all target photovoltaic panels are cleaned.

[0064] Example 2 The photovoltaic cleaning robot cleaning system, which applies the photovoltaic cleaning path and water volume scheduling method based on turbidity feedback, includes: a data acquisition unit, a soiling assessment unit, a mapping table management and water usage budget unit, a batch path planning unit, an execution control and turbidity acquisition unit, a compliance assessment unit, a water usage margin calculation and re-cleaning scheduling unit, a scheduling and result output unit, and a data processing module. The data acquisition unit, pollution assessment unit, mapping table management and water budgeting unit, batch path planning unit, execution control and turbidity acquisition unit, compliance assessment unit, water margin calculation and re-cleaning scheduling unit, and scheduling and result output unit are respectively connected to the data processing module. The data acquisition unit is used to acquire the board surface image, the board surface optical signal, and the ambient light intensity; The aforementioned soiling assessment unit is used to extract soiling features and generate soiling degree values ​​and stain type labels to form a board-level soiling list; The mapping table management and water use budget unit is used to determine the target water use at the board level from the preset mapping table, and generate a batch water use budget table and a board level water use allocation table, which include the allocated water use and the reserved water use, with the water tank capacity as the upper limit of water use in a single batch. The batch path planning unit is used to set the water replenishment location as the start and end point of the batch operation path and generate the batch operation path. The execution control and turbidity acquisition unit is used to perform spraying and scrubbing along the batch operation path, control the cleaning according to the plate level water allocation table, collect the turbidity of the return water in real time, and record the actual water consumption and operation time. The aforementioned compliance assessment unit is used to conduct compliance assessment based on the turbidity of the return water and the turbidity threshold of the return water, and output a candidate set for batch re-cleaning. The water margin calculation and re-cleaning scheduling unit is used to calculate the water margin of this batch based on the batch re-cleaning candidate set, the batch water budget table and the cumulative water consumption of this batch, generate a panel-level water allocation table for non-compliant photovoltaic panels in combination with the water consumption impact degree, and insert the re-cleaning path before the end of the batch operation path. The scheduling and result output unit is used to trigger water replenishment after the current batch ends and repeat steps one to five for the remaining batches, while outputting the batch water usage ledger, the result of the return water turbidity meeting the standard, and the re-cleaning record.

[0065] Example 3: Path-Dosage-Turbidity Closed Loop for Single Batch Operations Scenario: A section of a power plant has a full water tank, and the goal is to clean several rows of components within one water replenishment cycle.

[0066] Once the cleaning robot arrives at the work area, the data acquisition unit obtains images of the panel surface, optical signals from the panel surface, and ambient light intensity. The soiling assessment unit outputs a soiling level value and a stain type label for each panel based on preset feature rules, forming a panel-level soiling list.

[0067] The mapping table management and water budget unit reads the spraying and scrubbing parameters from the preset mapping table based on the degree of soiling / stain type to form the target water consumption at the board level; and generates a batch water budget table (allocation / reservation) and a board-level water allocation table with the water tank capacity as the upper limit of water consumption in a single batch.

[0068] The batch path planning unit sets the water replenishment location as the start and end point to generate batch operation paths.

[0069] The execution control and turbidity acquisition unit sweeps along the batch operation path, controls pumps and valves according to the plate-level water allocation table, collects and marks the turbidity of the return water online, records the actual water consumption and operation time, and calculates the cumulative water consumption of this batch on a rolling basis.

[0070] The compliance assessment unit compares the turbidity of the return water with the turbidity threshold. If the standard is met, the process is stopped in advance, and the remaining water consumption corresponding to the target is replenished to the reserved remaining water consumption and the ledger is updated; if the standard is not met, the water is added to the batch re-cleaning candidate set.

[0071] The water margin calculation and re-cleaning scheduling unit calculates the water impact and water demand of each board based on the candidate set, and generates a board-level water allocation table for boards that do not meet the standards after comparing it with the water margin of this batch. Before the end of the batch, a re-cleaning path is inserted and executed according to the principle of minimum incremental path. After completion, the batch ends and the batch water usage ledger, compliance results and re-cleaning records are output.

[0072] Example 4: Multi-batch scheduling and legacy list processing Scenario: In a large-scale power plant area, the capacity of a single water tank is insufficient to cover the entire panel, requiring multiple batches of circulation; during this period, there are substandard panels with insufficient water sufficiency and inefficient water replenishment.

[0073] As in Example 3, a panel-level contamination list and target water consumption are obtained. Due to the large coverage area, the system automatically generates multiple job batches based on the maximum water consumption per batch. These are executed sequentially according to the batch number: each batch completes path planning, execution, turbidity assessment, re-cleaning insertion, and batch termination.

[0074] In a certain batch, some non-compliant boards are marked as inefficient in terms of water use impact assessment. The system automatically reduces their board-level water demand or temporarily suspends their re-cleaning. When the water use margin for this batch is insufficient, the scheduling strategy eliminates candidates according to the water use impact from small to large and adds the eliminated boards to the legacy list.

[0075] After a batch is completed, the scheduling and result output unit triggers water replenishment, and the process moves to the next batch. The remaining water consumption of the target panels that have met the standards continues to be replenished to the new reserved water consumption, providing flexibility for subsequent batches of cleaning.

[0076] Once all batches are completed, the remaining list will be handled uniformly: manual centralized cleaning can be selected, or a dedicated re-cleaning batch can be created to independently plan and execute the re-cleaning path for the remaining boards.

[0077] In multi-batch scenarios, this solution ensures that the goal of effective water increase is achieved first under limited resource conditions by using the dual constraints of water use margin and water use impact for each batch. For objects with inefficient water increase or exceeding the upper limit, an auditable postponement strategy is provided to achieve a balance between overall water efficiency and compliance rate.

[0078] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0079] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0080] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A photovoltaic cleaning path and water volume scheduling method based on turbidity feedback, characterized in that, Includes the following steps: Step 1: Collect photovoltaic panel surface contamination information and generate a panel-level contamination list; generate a batch water usage budget table and a panel-level water usage allocation table based on the panel-level contamination list, and generate a batch operation path based on the batch water usage budget table and the panel-level water usage allocation table; During the cleaning operation of each photovoltaic panel along the batch operation path, the turbidity of the return water of each photovoltaic panel is collected in real time, and the actual water consumption and operation time of each photovoltaic panel are recorded. Step 2: Compare the turbidity of the return water from each photovoltaic panel with a preset return water turbidity threshold. When the turbidity of the return water from a photovoltaic panel is less than or equal to the threshold, stop cleaning that panel and determine that it has met the cleaning standard. Add the target remaining water volume of that photovoltaic panel to the reserved water volume in the batch water budget table and update the batch remaining water volume in the batch water budget table. When the turbidity of the return water is still not less than the threshold after the corresponding spraying time has been reached, stop cleaning that photovoltaic panel, determine that it has not met the cleaning standard, and add the corresponding photovoltaic panel to the batch re-cleaning candidate set. The target remaining water volume is the difference between the target water volume for each panel and its actual water volume. When the difference is less than zero, the target remaining water volume is zero. Step 3: Obtain the turbidity difference between the turbidity of the return water at the time of stopping cleaning for each non-compliant photovoltaic panel and the turbidity threshold value of the return water. Obtain the turbidity reduction value between the maximum value of the return water turbidity during the cleaning process and the turbidity of the return water at the time of stopping cleaning for each non-compliant photovoltaic panel. Based on the turbidity difference and the turbidity reduction value, and using the larger of the turbidity difference and the preset minimum difference value as the denominator, calculate the ratio of the turbidity reduction value to the denominator as the water use impact degree of the corresponding photovoltaic panel. Based on the product of the panel-level target water consumption and the water use impact degree of each non-compliant photovoltaic panel, obtain the panel-level water demand of each non-compliant photovoltaic panel. Step 4: Calculate the water margin for this batch based on the updated remaining water volume in the batch water budget table, and compare the sum of the water demand for each non-compliant photovoltaic panel with the remaining water volume of the batch: when the sum is less than the remaining water volume of the batch, the water demand for each non-compliant photovoltaic panel is directly used as the basis for generating the water allocation table for non-compliant photovoltaic panels; when the sum is not less than the remaining water volume of the batch, the corresponding non-compliant photovoltaic panels are removed in order of increasing water impact until the sum of the water demand for the remaining non-compliant photovoltaic panels is less than the remaining water volume of the batch, and then the water allocation table for the non-compliant photovoltaic panels is generated based on the remaining non-compliant photovoltaic panels. Step 5: Based on the panel-level water allocation table of the substandard photovoltaic panels, insert a re-cleaning path before the end of the batch operation path; after the re-cleaning path is completed, end the current batch, replenish water, and clean the next batch until the operation task is completed.

2. The photovoltaic cleaning path and water volume scheduling method based on turbidity feedback according to claim 1, characterized in that, The process involves collecting photovoltaic panel surface contamination information to obtain a panel-level contamination list, generating a batch water usage budget table and a panel-level water usage allocation table based on the panel-level contamination list, and obtaining a batch operation path based on the batch water usage budget table and the panel-level water usage allocation table, including: The system collects images of the photovoltaic panel surface, optical signals from the panel surface, and ambient light intensity, extracts dirt features, generates dirt severity values ​​and stain type labels, and forms a panel-level dirt list. Based on the panel-level dirt list, the system determines the target water consumption for each photovoltaic panel from a preset mapping table. The system uses the water tank capacity of the cleaning robot as the upper limit of water consumption per batch to divide the work tasks into batches, generating a batch water consumption budget table and a panel-level water consumption allocation table. Based on the panel-level water consumption allocation table, the water replenishment location is set as the start and end points of the batch operation path, generating the batch operation path.

3. The photovoltaic cleaning path and water volume scheduling method based on turbidity feedback according to claim 2, characterized in that, The process involves acquiring images of the board surface, optical signals from the board surface, and ambient light intensity; extracting soiling features; generating soiling severity values ​​and stain type labels; and forming a board-level soiling list, including: The photovoltaic panel data acquisition module collects image data and optical signals of the photovoltaic panel surface. The stain type is identified based on the set stain type identification features and the image data. The optical signal is the reflected light intensity obtained when light of a preset intensity shines on the photovoltaic panel. Based on a pre-calibrated reflection intensity-stain degree mapping relationship, the stain degree value is obtained according to the reflection intensity. The stain degree value and the stain type are used to form a panel-level stain list.

4. The photovoltaic cleaning path and water volume scheduling method based on turbidity feedback according to claim 2, characterized in that, Based on the panel-level contamination list, the target water consumption for each photovoltaic panel is determined from a preset mapping table. The work tasks are divided into batches using the water tank capacity as the upper limit for water consumption in a single batch. A batch water consumption budget table and a panel-level water consumption allocation table are generated, including: Read the nozzle opening corresponding to the stain type and the spray duration corresponding to the degree of staining from the preset mapping table, and determine the target water consumption at the plate level based on the spray duration and the nozzle opening; The cleaning robot's water tank capacity is used as the upper limit for water consumption per batch to divide the work tasks into batches, forming a batch water consumption budget table that includes allocated water consumption and reserved water consumption. The reserved water consumption is used for re-cleaning and abnormal consumption in this batch. Based on the allocated water consumption and the target water consumption for each panel, the photovoltaic panels that can be cleaned in this batch are determined, and a panel-level water consumption allocation table is formed based on the information of the cleanable photovoltaic panels, the spraying duration, the nozzle opening, and the target water consumption for each panel. The sum of the allocated water consumption and the reserved water consumption is equal to the water tank capacity. The panel-level water consumption allocation table corresponds one-to-one with the spraying duration and nozzle opening parameters and the target water consumption for each panel.

5. The photovoltaic cleaning path and water volume scheduling method based on turbidity feedback according to claim 3, characterized in that, The step of setting the water replenishment location as the start and end point of the batch operation path based on the plate-level water allocation table, and generating the batch operation path, includes: Obtain the location of each photovoltaic panel in the panel-level water allocation table to obtain the arrangement order of the photovoltaic panel locations. Add water replenishment locations at the beginning and end of the arrangement order of the photovoltaic panel locations to obtain the batch operation node order. Generate the batch operation path according to the batch operation node order.

6. The photovoltaic cleaning path and water volume scheduling method based on turbidity feedback according to claim 1, characterized in that, The step of calculating the water leeway for this batch based on the updated remaining water volume in the batch water budget table includes: Calculate the total water demand of each non-compliant photovoltaic panel; if the total is less than the remaining water demand of the batch, the water demand of this batch is determined to be sufficient; otherwise, the water demand of this batch is determined to be insufficient.

7. The photovoltaic cleaning path and water volume scheduling method based on turbidity feedback according to claim 1, characterized in that, The panel-level water allocation table based on substandard photovoltaic panels inserts a re-cleaning path before the end of the batch operation path, including: The sequence of operation nodes for the substandard photovoltaic panels is determined based on the panel-level water allocation table, a re-cleaning path is generated, and the re-cleaning path is inserted before the end point of the batch operation path.

8. The photovoltaic cleaning path and water volume scheduling method based on turbidity feedback according to claim 1, characterized in that, The step of ending the current batch after completing the re-cleaning path also includes: The non-compliant photovoltaic panels that were rejected in each batch were compiled into a legacy list. The photovoltaic panels in the legacy list were either manually processed in a unified manner or cleaned separately.

9. A photovoltaic cleaning robot cleaning system, characterized in that, The photovoltaic cleaning path and water volume scheduling method based on turbidity feedback as described in any one of claims 1-8 includes: a data acquisition unit, a pollution assessment unit, a mapping table management and water use budget unit, a batch path planning unit, an execution control and turbidity acquisition unit, a compliance assessment unit, a water use margin calculation and re-cleaning scheduling unit, a scheduling and result output unit, and a data processing module. The data acquisition unit, pollution assessment unit, mapping table management and water budgeting unit, batch path planning unit, execution control and turbidity acquisition unit, compliance assessment unit, water margin calculation and re-cleaning scheduling unit, and scheduling and result output unit are respectively connected to the data processing module. The data acquisition unit is used to acquire the board surface image, board surface optical signal and ambient light intensity; The soiling assessment unit is used to extract soiling features and generate soiling degree values ​​and stain type labels to form a board-level soiling list. The mapping table management and water use budget unit is used to determine the target water consumption at the board level from the preset mapping table, and generate a batch water use budget table and a board level water use allocation table, which include the allocated water consumption and the reserved water consumption, with the water tank capacity as the upper limit of water consumption in a single batch. The batch path planning unit is used to set the water replenishment location as the start and end point of the batch operation path and generate the batch operation path. The execution control and turbidity acquisition unit is used to perform spraying and brushing along the batch operation path, control the cleaning according to the plate water allocation table, collect the turbidity of the return water in real time, and record the actual water consumption and operation time. The compliance assessment unit is used to conduct compliance assessment based on the turbidity of the return water and the turbidity threshold of the return water, and output a batch re-cleaning candidate set. The water margin calculation and re-cleaning scheduling unit is used to calculate the water margin of this batch based on the batch re-cleaning candidate set, the batch water budget table and the cumulative water consumption of this batch, generate the panel-level water allocation table for non-compliant photovoltaic panels in combination with the water impact degree, and insert the re-cleaning path before the end of the batch operation path. The scheduling and result output unit is used to trigger water replenishment after the current batch ends and continue cleaning the remaining batches, while outputting the batch water usage log, the result of the return water turbidity meeting the standard, and the re-cleaning record.

Images