Wireless sensor network coverage scheduling method and system for energy collection

Abstract

The invention belongs to the technical field of wireless sensor network application, and discloses a wireless sensor network coverage scheduling method and system oriented to energy harvesting, and the method comprises the steps: constructing an intra-day multi-scene power track according to the historical weather and photovoltaic data collected by each sensor, introducing tail risk measurement to evaluate the track, and carrying out the prediction of the track. Outputting a risk-controllable intra-day prediction power curve, and calculating the current-day energy budget of each sensor; dividing an intra-day time slot into different energy balance periods according to an intra-day prediction power curve, and establishing available energy constraints in the balance periods; gridding a preset monitoring area, and evaluating the collaborative detectable potential of a grid in combination with the energy budget of each sensor; grids meeting continuity constraints are selected from the preset space column structure, and a continuous target grid chain is formed; and the coverage quality, the energy utilization rate and the sustainability of the wireless sensor network are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of wireless sensor network application technology, and more specifically, to a wireless sensor network coverage scheduling method and system for energy harvesting. Background Technology

[0002] Existing wireless sensor network coverage scheduling methods and systems mainly suffer from the following problems: Wireless sensor networks have attracted significant research and application attention due to their wide range of applications across various fields. In practical deployments, sensor nodes are powered by limited batteries, making network operation energy-constrained. In recent years, energy-harvesting wireless sensor networks have become a research hotspot, as the harvesting of renewable energy sources such as photovoltaics can alleviate the energy limitations of nodes. However, existing wireless sensor network coverage scheduling methods and systems have many shortcomings.

[0003] Existing methods only provide a single power prediction curve, failing to consider photovoltaic power deviations under extreme weather conditions. This makes it difficult to quantify tail risks, leading to energy gaps or curtailment in practical applications and reducing the reliability of network coverage scheduling. Secondly, traditional node energy budgeting relies on historical average power or empirical values, ignoring dynamic changes and fluctuations in photovoltaic power, resulting in unreasonable energy allocation and difficulty in ensuring continuous node operation.

[0004] Furthermore, existing technologies lack methods for quantifying wasted solar energy, making it impossible to effectively control it. Traditional methods do not consider the differences in energy consumption across different node operating modes, leading to network coverage issues. Existing methods rely on experience or static models, lacking consideration of node collaboration, thus rendering coverage optimization lacking a scientific basis. Simultaneously, the failure to dynamically select grid cells based on feasible sensing radii prevents adaptive adjustment of coverage scheduling, reducing the long-term reliability of the network.

[0005] In view of this, the present invention proposes a wireless sensor network coverage scheduling method for energy harvesting to solve the above problems. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a wireless sensor network coverage scheduling method for energy harvesting, comprising: S1. Based on the historical meteorological and photovoltaic data collected by each sensor, construct the intraday multi-scenario power trajectory, introduce tail risk measurement to evaluate the trajectory, output the intraday predicted power curve with controllable risk, and calculate the daily energy budget of each sensor. S2. Based on the intraday predicted power curve, the intraday time slot is divided into different energy balance cycles, and available energy constraints are established within the balance cycle. S3. Grid the preset monitoring area, evaluate the collaborative detectability potential of the grid by combining the energy budget of each sensor, and select the grid that meets the continuity constraint in the preset spatial column structure to form a continuous target grid chain. S4. Calculate the cooperative coverage quality of each grid in the target grid chain in each time slot, search for the grid-time combination with the smallest cooperative coverage quality within the range of the target grid chain and time slot combination, and identify the spatiotemporal bottleneck point with the weakest coverage. S5. Under the constraint of available energy, with the collaborative coverage quality of spatiotemporal bottlenecks as the primary objective, construct a work schedule for each sensor; repair the schedule that exceeds the energy limit, and perform fusion scoring on the schedule under different energy budgets to obtain the optimal schedule; each sensor executes monitoring tasks according to the optimal work schedule and reports its operating status.

[0007] Preferably, the method for constructing intraday multi-scenario power trajectories includes: Continuous data collection is achieved through meteorological sensors and photovoltaic power output monitoring units deployed in the monitoring area. The meteorological sensors are used to acquire meteorological data, and the photovoltaic power output monitoring units are used to acquire photovoltaic power output data at the corresponding time. A unified time identifier is assigned to both the meteorological data and the photovoltaic power output data. The meteorological data and photovoltaic output power data are processed by timestamp alignment, outlier removal, missing value imputation and unified sampling interval to form a one-to-one corresponding sample sequence at the same time scale; and the meteorological data and photovoltaic output power data are processed by standard deviation normalization to construct a meteorological state index characterizing the intensity of the comprehensive meteorological effect, and the changing trend of the meteorological state index is extracted. The normalized meteorological state indicators and their changing trends are used as historical samples. The mean, variance, and quantiles of the historical samples in terms of indicator values ​​and rates of change are statistically analyzed. The indicator values ​​refer to the meteorological state indicators at a single moment, and the rates of change refer to the rates of change of the meteorological state indicators between consecutive time slots. By representing each historical sample as a sample vector, which includes the meteorological state index value of the sample in each time slot and the corresponding rate of change of the meteorological state index, the cosine similarity between any two sample vectors is calculated. If the cosine similarity of two sample vectors is greater than or equal to the preset cosine similarity threshold, the two samples are determined to have similar meteorological action mechanisms and photovoltaic power response characteristics. Based on the cosine similarity calculation results, all historical samples are divided into different meteorological scene types. Each meteorological scene type corresponds to a set of similar samples, and scene identifiers and frequency weights are assigned. For each meteorological field scenario type, combined with the theoretical irradiance law corresponding to the intraday solar altitude angle change, a theoretical power output envelope curve under the condition of no meteorological disturbance is constructed, and an intraday power trajectory is generated based on the power change characteristics of historical samples. The trajectories of each scenario are combined in chronological order and branched according to the frequency of occurrence to form a multi-scenario photovoltaic power scenario tree within a day. Each branch represents a possible power evolution path, thus forming a multi-scenario photovoltaic power trajectory within a day.

[0008] Preferably, the method for calculating the daily energy budget for each sensor includes: For each candidate power trajectory in the multi-scenario photovoltaic power trajectory within the day, a tail risk metric is introduced to evaluate the candidate power trajectory, and a comprehensive evaluation function is formed by combining energy utilization utility and curtailment loss. The risk perception score of each candidate power trajectory is then obtained through the comprehensive evaluation function. Based on the risk perception score, a candidate power trajectory with the highest energy utilization efficiency, the lowest curtailment loss, and controllable tail risk is selected to generate an intraday predicted power curve. Based on the intraday predicted power curve, the power value corresponding to each sensor is obtained in each time slot, and combined with the rated capacity of each sensor, the daily energy budget of each sensor is calculated.

[0009] Preferably, the method for establishing an available energy constraint within the equilibrium period includes: The intraday predicted power curve is divided into different continuous time slots according to the time series, and each time slot corresponds to a fixed time interval. Based on the intraday predicted power curve, adjacent time slots are divided into different energy balance periods according to the cumulative available energy of the sensor, so that the total available energy in each balance period meets the sum of the predicted power demand in that energy balance period. Within each energy balance cycle, available energy constraints are established for each sensor. These constraints include that the total energy consumption of the sensor within the energy balance cycle is less than or equal to the available energy budget of the sensor within that energy balance cycle, and that the power consumption of the sensor within each time slot is less than or equal to the predicted power of the sensor corresponding to that time slot, thus forming a scheduling time organization framework.

[0010] Preferably, the method for assessing the collaborative detectability potential of the budget assessment grid includes: The preset monitoring area is divided into different grid units according to the preset spatial resolution; based on the spatial positional relationship between the sensor and the grid unit and the preset probability perception model, the contribution of each sensor to the single-node detection of the grid unit at different preset perception radius levels is calculated. Obtain the daily energy budget and the corresponding working energy consumption of each sensing radius level for each sensor; for each grid cell, select feasible radius levels from the set of selectable sensing radius levels for each sensor that satisfy the requirement that the daily energy budget is not less than the energy consumption of that sensing radius level; Then, the radius level that makes the largest contribution to the detection of a single node in the grid cell is selected from the feasible radius levels as the effective detection contribution of the sensor to the grid cell; the effective detection contributions of all sensors are accumulated to obtain the cooperative detectable potential of the grid cell under energy constraints.

[0011] Preferably, the method for forming a continuous target mesh chain includes: For each grid cell, grid cells that meet the energy feasibility criteria are selected as candidate grid cells based on the cooperative detectability potential of each grid cell under energy constraints and the minimum energy required for each sensor to maintain operation on each grid cell. One target grid is selected sequentially from the candidate grid cells in each column. During the selection process, the spatial distance between the selected target grids in adjacent columns is less than or equal to a preset spatial distance threshold. At the same time, grids with cooperative detectability potential greater than the preset cooperative detectability potential threshold are selected first, thereby obtaining a continuous grid sequence from the first column to the last column, which forms a continuous target grid chain.

[0012] Preferably, the method for identifying the spatiotemporal bottleneck point with the weakest coverage includes: For each grid in the formed continuous target grid chain, the single-node detection contributions of each sensor are fused according to the preset collaborative fusion rules in each time slot of the prediction day to obtain the collaborative coverage quality of the grid in that time slot. Within the range of continuous target grid chains and time slots, the search collaboratively covers the grid and time point with the lowest quality as a grid-time combination, and this grid-time combination covers the weakest spatiotemporal bottleneck point.

[0013] Preferably, the method for obtaining the optimal schedule includes: A scheduling search space is established for each sensor. The working status of the sensor in each time slot of the prediction day and the corresponding sensing radius level are combined and encoded as candidate scheduling individuals. A local scheduling population is maintained for each sensor. The candidate scheduling individuals are subjected to evolutionary search by iteratively performing selection, crossover, mutation and elite retention operations. During the evolution process, the improvement of collaborative coverage quality at the spatiotemporal bottleneck point is used as the optimization objective. Under the premise of ensuring that the optimization objective is not reduced, the collaborative coverage quality of the continuous target grid chain is further optimized, thus forming a lexicographical dual-objective optimization mechanism to obtain the local optimal scheduling of a single sensor. For candidate scheduling individuals whose energy consumption exceeds the available energy budget within the energy balance cycle during the evolution process, the candidate scheduling individuals with the lowest energy cost-effectiveness are gradually reduced or their work level is lowered until the available energy budget is met, based on the unit energy coverage gain corresponding to each time slot and each perception radius level, so as to achieve scheduling feasibility repair. During the evolution of each sensor, elite scheduling summary information is exchanged at low frequency. When the evolution reaches the preset termination condition, the local optimal schedules obtained under different energy budget conditions are fused and scored to obtain the comprehensive score of the individual schedule. Finally, the local optimal schedule with the highest comprehensive score of the individual schedule is selected as the optimal schedule of each sensor.

[0014] Preferably, the method for each sensor to perform monitoring tasks and report its operating status according to the optimal work schedule includes: Each sensor performs monitoring tasks according to the optimal schedule obtained, within each time slot of the day, based on the sensing radius level determined by the schedule. There are preset working time slots and sleeping time slots. During the working time slot, the sensor collects monitoring data and detects targets in the grid cells covered by the continuous target grid chain. During the sleeping time slot, the sensor enters standby mode, and at the same time, the sensor reports the collected monitoring data and its own operating status to the network control terminal through a preset communication strategy. A wireless sensor network coverage scheduling system for energy harvesting includes: The risk perception and prediction module is used to construct intraday multi-scenario power trajectories based on historical meteorological and photovoltaic data collected by various sensors, introduce tail risk measurement to evaluate the trajectories, output intraday predicted power curves with controllable risks, and calculate the daily energy budget for each sensor. The energy balance construction module is used to divide the intraday time slot into different energy balance periods based on the intraday predicted power curve, and to establish available energy constraints within the balance period. The target chain selection module is used to grid the preset monitoring area, evaluate the collaborative detectability potential of the grid by combining the energy budget of each sensor, and select the grid that meets the continuity constraint in the preset spatial column structure to form a continuous target grid chain. The spatiotemporal bottleneck identification module is used to calculate the cooperative coverage quality of each grid in the target grid chain in each time slot, search for the grid-time combination with the smallest cooperative coverage quality within the range of the target grid chain and time slot combination, and identify the spatiotemporal bottleneck point with the weakest coverage. The distributed scheduling module is used to construct work schedules for each sensor under the constraint of available energy, with the collaborative coverage quality of spatiotemporal bottlenecks as the primary objective; it also repairs schedules that exceed energy limits, and performs fusion scoring on schedules under different energy budgets to obtain the optimal schedule; each sensor executes monitoring tasks according to the optimal work schedule and reports its operating status.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This invention evaluates intraday photovoltaic power trajectories across multiple scenarios by introducing a tail risk metric. This quantifies power deviations under extremely unfavorable weather conditions, generates predictive power curves with controllable risks, and reduces power shortfalls and coverage interruption risks. By comprehensively considering energy utilization efficiency and curtailment losses, it achieves efficient utilization of photovoltaic power generation resources and reduces energy waste. Tail risk perception scoring is applied to photovoltaic power trajectories across multiple scenarios, and the generated predictive power curves take into account power fluctuations under different weather conditions, providing dynamic and quantifiable input for wireless sensor network coverage scheduling.

[0016] While ensuring sufficient sensor energy, the work schedule is dynamically adjusted through a risk-controlled energy budget and power prediction to balance coverage quality and energy efficiency, thereby improving the long-term operational performance of the network. Feasible sensing radius levels that meet the daily energy budget are selected, and the single node with the largest detection contribution is chosen as the effective contribution, ensuring optimal sensor coverage and avoiding energy waste or coverage interruptions. The collaborative detectable potential of grid cells with effective detection contributions from all sensors is accumulated, providing a quantitative basis for coverage scheduling and helping to identify weak areas and optimize scheduling. The collaborative detectable potential of the grid under different energy conditions is dynamically evaluated, allowing for flexible adjustment of sensor operating modes to balance coverage quality and energy consumption, improving the long-term stability of the network. The quantified results are used to construct continuous target grid chains, selecting the path with the maximum coverage potential under spatial continuity constraints, improving coverage effectiveness and task completion rate in the monitored area. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the wireless sensor network coverage scheduling method for energy harvesting according to the present invention; Figure 2 This is a schematic diagram of the structure of the wireless sensor network coverage scheduling system for energy harvesting according to the present invention. Detailed Implementation

[0018] 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. Example 1

[0019] Please see Figure 1 As shown, this embodiment provides a wireless sensor network coverage scheduling method for energy harvesting, specifically including the following steps: S1. Based on the historical meteorological and photovoltaic data collected by each sensor, construct the intraday multi-scenario power trajectory, introduce tail risk measurement to evaluate the trajectory, output the intraday predicted power curve with controllable risk, and calculate the daily energy budget of each sensor. S2. Based on the intraday predicted power curve, the intraday time slot is divided into different energy balance cycles, and available energy constraints are established within the balance cycle. S3. Grid the preset monitoring area, evaluate the collaborative detectability potential of the grid by combining the energy budget of each sensor, and select the grid that meets the continuity constraint in the preset spatial column structure to form a continuous target grid chain. S4. Calculate the cooperative coverage quality of each grid in the target grid chain in each time slot, search for the grid-time combination with the smallest cooperative coverage quality within the range of the target grid chain and time slot combination, and identify the spatiotemporal bottleneck point with the weakest coverage. S5. Under the constraint of available energy, with the collaborative coverage quality of spatiotemporal bottlenecks as the primary objective, construct a work schedule for each sensor; repair the schedule that exceeds the energy limit, and perform fusion scoring on the schedule under different energy budgets to obtain the optimal schedule; each sensor executes monitoring tasks according to the optimal work schedule and reports its operating status.

[0020] Methods for constructing intraday multi-scenario power trajectories include: The system continuously collects data through meteorological sensors and photovoltaic power output monitoring units deployed in the monitoring area. The meteorological sensors are used to acquire meteorological data, including irradiance, ambient temperature, module temperature, humidity, wind speed, and cloud cover. The photovoltaic power output monitoring units are used to acquire photovoltaic output power data at the corresponding time and assign a unified time identifier to the meteorological data and photovoltaic output power data. The meteorological sensors include irradiance sensors, ambient temperature sensors, humidity sensors, wind speed sensors, and temperature sensors that acquire the operating temperature of the components; the photovoltaic output monitoring unit includes voltage sensors and current sensors. The meteorological data and photovoltaic output power data are processed by timestamp alignment, outlier removal, missing value imputation and unified sampling interval to form a one-to-one corresponding sample sequence at the same time scale; and the meteorological data and photovoltaic output power data are processed by standard deviation normalization to construct a meteorological state index characterizing the intensity of the comprehensive meteorological effect, and the changing trend of the meteorological state index is extracted. It should be noted that timestamp alignment is based on a unified time synchronization mechanism, with each acquisition unit synchronizing its clock through a network time protocol or satellite time synchronization. When there is still a millisecond-level deviation between different devices, the non-aligned data is mapped to the standard time point using linear interpolation or hold-type resampling with reference to a preset benchmark time axis.

[0021] Outlier removal is generally achieved based on dual constraints of physical feasible interval and statistical deviation. For example, data with irradiance less than zero or exceeding the upper limit of the theoretical solar constant are directly judged as invalid. At the same time, the moving window mean and standard deviation of the power series are calculated, and data exceeding the mean plus or minus k times the standard deviation are removed as outliers.

[0022] Missing value imputation can be performed by linear interpolation of adjacent valid points when the missing value is short-term, and by using historical averages of the same meteorological scene or typical curves of the same time period when the missing value is continuous for a long time. Unifying the sampling interval is usually achieved through resampling techniques, that is, with a fixed time step as the target, high-frequency data is downsampled by averaging, and low-frequency data is upsampled by interpolation, so as to form a sequence with the same time step.

[0023] Meteorological sensors deployed in the monitoring area collect information on irradiance, ambient temperature, module temperature, relative humidity, wind speed, and cloud cover. These data are then normalized to eliminate differences in dimensions and magnitudes. The normalized meteorological quantities are then combined linearly or nonlinearly according to preset weights to form a single meteorological state index, used to quantify the overall driving force of meteorological conditions on photovoltaic power output. Subsequently, the changing trends of the meteorological state index over continuous time series are extracted. This can be achieved by calculating the local average rate of change using a sliding window, analyzing the short-term direction and rate of change using first-order difference or linear regression methods, or extracting intraday trends through low-pass filtering, thereby obtaining the trend information of meteorological state evolution over time.

[0024] The normalized meteorological state indicators and their changing trends are used as historical samples. The mean, variance, and quantiles of the historical samples in terms of indicator values ​​and rates of change are statistically analyzed. The indicator values ​​refer to the meteorological state indicators at a single moment, and the rates of change refer to the rates of change of the meteorological state indicators between consecutive time slots. By representing each historical sample as a sample vector, which includes the meteorological state index value of the sample in each time slot and the corresponding rate of change of the meteorological state index, the cosine similarity between any two sample vectors is calculated. If the cosine similarity of two sample vectors is greater than or equal to the preset cosine similarity threshold, the two samples are determined to have similar meteorological action mechanisms and photovoltaic power response characteristics. Based on the cosine similarity calculation results, all historical samples are divided into different meteorological scene types. Each meteorological scene type corresponds to a set of similar samples, and scene identifiers and frequency weights are assigned. For each meteorological field scenario type, combined with the theoretical irradiance law corresponding to the intraday solar altitude angle change, a theoretical power output envelope curve under the condition of no meteorological disturbance is constructed, and an intraday power trajectory is generated based on the power change characteristics of historical samples. It should be noted that, for each meteorological scene type, the solar altitude angle information corresponding to each time slot within the day is first obtained. Then, based on the relationship between the solar altitude angle and geographical location, date, and time, the theoretical irradiance is calculated. The theoretical irradiance represents the ideal energy input of solar radiation irradiating the surface of the photovoltaic module under conditions of no meteorological disturbance. Subsequently, based on the rated power of the photovoltaic module and the temperature correction coefficient, the theoretical irradiance is converted into the theoretical output power of the module, forming a theoretical power output sequence for the continuous time slots within the day.

[0025] Based on this, statistical analysis is performed on the historical sample power sequences assigned to the current meteorological field, including characteristic parameters such as power rise rate, fall rate, and fluctuation amplitude. According to the power change patterns of the historical samples, a range of possible power change values ​​within each time slot is generated. This power change range is limited to the envelope of the aforementioned theoretical output power to ensure the physical feasibility of the generated power values. By overlaying the power change characteristics of historical samples time-slot by time slot, combined with the theoretical output envelope constraint, an intraday power trajectory corresponding to the current meteorological field type can be generated. This power trajectory reflects the possible evolution path of photovoltaic power output over time under a specific meteorological scenario.

[0026] The trajectories of each scenario are combined in chronological order and branched according to the frequency of occurrence to form a multi-scenario photovoltaic power scenario tree within a day. Each branch represents a possible power evolution path, thus forming a multi-scenario photovoltaic power trajectory within a day.

[0027] The methods for calculating the daily energy budget for each sensor include: For each candidate power trajectory in the multi-scenario photovoltaic power trajectory within the day, a tail risk metric is introduced to evaluate the candidate power trajectory, and a comprehensive evaluation function is formed by combining energy utilization utility and curtailment loss. The risk perception score of each candidate power trajectory is then obtained through the comprehensive evaluation function. The comprehensive evaluation function is: ;in, This represents the risk perception score of the candidate power trajectory, used to compare the merits of different trajectories; Weighting coefficients representing energy utilization are used to adjust power utilization utility. Its importance in comprehensive evaluation; Energy utilization efficiency represents the total amount of energy effectively utilized by the photovoltaic power trajectory within the predicted day, which is calculated by integrating the predicted power. The weighting coefficient for tail risk is used to adjust the importance of loss risk in the overall evaluation under extremely adverse conditions. It represents conditional value at risk, a tail risk metric used to quantify the potential loss of a candidate power trajectory under the worst-case scenario; This indicates the level of risk confidence and controls the sensitivity of tail risk measures, such as 5% or 10%. Represents the energy gap random variable, representing the loss distribution or deviation from the target power; The weighting coefficient representing the curtailment loss is used to adjust for curtailment loss caused by excessively high power forecasts. Impact on overall evaluation; This represents the energy wasted due to unused photovoltaic power generation. Based on the risk perception score, a candidate power trajectory with the highest energy utilization efficiency, the lowest curtailment loss, and controllable tail risk is selected to generate an intraday predicted power curve. Based on the intraday predicted power curve, the power value corresponding to each sensor is obtained in each time slot, and combined with the rated capacity of each sensor, the daily energy budget of each sensor is calculated.

[0028] The daily energy budget is: ;in, Indicates sensor The energy budget for the forecast day, i.e., the total energy that the sensor can use or is expected to emit in a day; Indicates the sensor index; Indicates sensor In the The predicted power for each time slot is derived from the intraday predicted power curve; This indicates the duration of each time slot; Indicates the time slot index; This indicates the total number of time slots predicted for the day, determined by the time resolution. For example, if 24 hours are divided into 15-minute intervals, then... ; Methods for establishing usable energy constraints within the equilibrium period include: The intraday predicted power curve is divided into different continuous time slots according to the time series, and each time slot corresponds to a fixed time interval. Based on the intraday predicted power curve, adjacent time slots are divided into different energy balance periods according to the cumulative availability of sensor energy, so that the total available energy in each balance period meets the sum of the predicted power demand in that energy balance period, thereby ensuring that the sensor can complete the predetermined coverage task within the period. It should be noted that in this invention, the intraday predicted power curve is first discretized into a continuous time slot sequence according to a preset time resolution, and the predicted power value corresponding to each time slot is obtained. The predicted power is integrated according to the time slot duration to obtain the available energy value within each time slot, and then accumulated along the time axis to form a cumulative available energy sequence of the sensor changing with time throughout the day. At the same time, based on the coverage task's operational requirements for the sensor in each time slot, the predicted energy requirement for the corresponding time slot is determined, and similarly accumulated over time to obtain a cumulative energy requirement sequence.

[0029] Subsequently, the relationship between accumulated available energy and accumulated energy demand is compared sequentially along the time axis. When the accumulated available energy starting from a certain initial time slot first reaches or exceeds the accumulated energy demand for the corresponding time period, that time period is determined as the end point of an energy balance cycle, and the end time slot is used as the starting point of the next cycle to continue the judgment. In this way, the continuous time slots within the day are divided into several interconnected energy balance cycles, so that the total available energy in each energy balance cycle is not less than the sum of the predicted energy demand in that cycle, thereby ensuring that the sensor has the energy foundation to complete the corresponding coverage task on a cycle scale.

[0030] Within each energy balance cycle, available energy constraints are established for each sensor. These constraints include that the total energy consumption of the sensor within the energy balance cycle is less than or equal to the available energy budget of the sensor within that energy balance cycle, and that the power consumption of the sensor within each time slot is less than or equal to the predicted power of the sensor corresponding to that time slot, thus forming a scheduling time organization framework.

[0031] It should be noted that the process of obtaining the available energy budget for the sensor within the energy balance cycle is as follows: First, the intraday predicted power curve provides the predicted power value for each time slot in the form of discrete time slots, and the duration of each time slot is known; for any energy balance cycle, the index range of its start and end time slots is determined, the predicted power values ​​within this range are integrated and accumulated over time, and the predicted power of each time slot is multiplied by the corresponding time slot duration and summed to obtain the theoretical total energy available within the cycle; then, combined with the sensor's energy conversion efficiency, the energy storage device's charging and discharging efficiency, and power regulation losses, the theoretical total energy available is adjusted for efficiency to obtain the actual available energy value that can be used for sensor operation, and this value is used as the available energy budget for the sensor within the corresponding energy balance cycle.

[0032] Methods for budgeting the collaborative detectability potential of grids include: The preset monitoring area is divided into different grid units according to the preset spatial resolution; based on the spatial positional relationship between the sensor and the grid unit and the preset probability perception model, the contribution of each sensor to the single-node detection of the grid unit at different preset perception radius levels is calculated. It should be noted that, in this invention, based on the sensor Spatial location and grid cell The system determines the positional relationship between the two entities, calculates the Euclidean distance or spatial relative position parameters between them, and then uses a pre-defined probability sensing model to classify the distance, angle, and sensor radius levels. As input, the probability that the sensor will monitor the grid cell at that radius level is calculated, and the probability value is used as the sensor's single-node detection contribution to the grid cell. The probabilistic sensing model can adopt an empirical function, an exponential decay model, or a Gaussian decay model to reflect the law of sensor detection capability decay with distance. At the same time, it can be corrected by combining the sensor's photosensitivity, signal-to-noise ratio, and other physical characteristics. Through the above method, the single-node detection contribution of each sensor to each grid cell at different radius levels is obtained.

[0033] Sensing radius levels are discretized representations of the selectable sensing range of a sensor, used to characterize the relationship between the sensor's coverage capability and energy consumption in different operating modes. Specifically, each sensor defines several sensing radius levels based on its hardware performance, antenna characteristics, and energy consumption model. Each level corresponds to a specific sensing radius value and the operating energy required for monitoring at that radius. Sensing radius levels can be obtained through experimental calibration, theoretical model calculations, or performance parameters provided by the sensor manufacturer, including the minimum and maximum achievable radii of the sensor in different power modes, as well as several intermediate discrete levels. By discretizing the sensing radius into a set of levels, a quantitative description of the sensor's coverage capability can be achieved while ensuring energy controllability.

[0034] Obtain the daily energy budget and the corresponding working energy consumption of each sensing radius level for each sensor; for each grid cell, select feasible radius levels from the set of selectable sensing radius levels for each sensor that satisfy the requirement that the daily energy budget is not less than the energy consumption of that sensing radius level; Then, the radius level that makes the largest contribution to the detection of a single node in the grid cell is selected from the feasible radius levels as the effective detection contribution of the sensor to the grid cell; the effective detection contributions of all sensors are accumulated to obtain the cooperative detectable potential of the grid cell under energy constraints.

[0035] The potential for collaborative detection is: ; where represents the grid. The cooperative detection potential represents the upper bound of the cooperative detection capability that the grid theory can achieve under energy-feasible conditions. Represents a set of sensors; Indicates sensor A set of optional perception radius levels; Indicates sensor In radius level Below the grid The single-node detection contribution is calculated by the probabilistic perception model and spatial relationships; Indicates an indicator function, when The value is 1 if the condition is met, otherwise it is 0. Indicates the first Each sensor at the radius level The minimum energy required to maintain operation; Indicates the grid index; Methods for forming continuous target mesh chains include: For each grid cell, grid cells that meet the energy feasibility criteria are selected as candidate grid cells based on the cooperative detectability potential of each grid cell under energy constraints and the minimum energy required for each sensor to maintain operation on each grid cell. It should be noted that, in this invention, for each column of grid cells in the preset monitoring area, the system first obtains the cooperative detectability potential of each grid cell under energy constraints, that is, considering the detection capability that multiple sensors can achieve by cooperating in monitoring the grid cell under the limitation of available sensor energy. Simultaneously, the system obtains the minimum energy required for each sensor to maintain basic operation in each grid cell, that is, the minimum energy consumption required for the sensor to perform the monitoring task. Subsequently, the system performs an energy feasibility judgment on each grid cell: only when the cooperative detectability potential of the grid cell is achievable within the range of available sensor energy, and the available energy of each participating sensor is greater than or equal to the minimum energy required to maintain basic operation in that grid cell, is the grid cell determined to be an energy-feasible grid.

[0036] One target grid is selected sequentially from the candidate grid cells in each column. During the selection process, the spatial distance between the selected target grids in adjacent columns is less than or equal to a preset spatial distance threshold to ensure that the resulting grid sequence is spatially continuous or connectable. At the same time, grids with cooperative detection potential greater than a preset cooperative detection potential threshold are selected first, thereby obtaining a continuous grid sequence from the first column to the last column, i.e., forming a continuous target grid chain.

[0037] Methods for identifying the spatiotemporal bottlenecks with the weakest coverage include: For each grid in the formed continuous target grid chain, the single-node detection contributions of each sensor are fused according to the preset collaborative fusion rules in each time slot of the prediction day to obtain the collaborative coverage quality of the grid in that time slot. It should be noted that, in this invention, the fusion process of the collaborative fusion rule includes the following steps: First, based on the probabilistic perception model, the single-node detection contribution of each sensor to the target grid within the time slot is obtained. This contribution reflects the reliable detection capability of the sensor to the grid under energy constraints. Then, according to the reliability, historical detection success rate, energy availability, and preset task priority of each sensor, the single-node detection contribution is assigned a corresponding weight. Next, the weighted contributions of all sensors are combined according to a preset fusion strategy. The fusion strategy may include weighted summation, taking the maximum value, or probabilistic complementary logic operation to comprehensively reflect the overall detection capability of the entire sensor set to the grid. During the fusion process, for sensors whose energy is insufficient to support the current detection task, their single-node contribution will be set to zero or reduced proportionally. Finally, the fusion result is used as the collaborative coverage quality of the grid within the time slot, quantitatively describing the degree to which the grid is reliably detected by the sensor set within the time slot.

[0038] Within the range of continuous target grid chains and time slots, the search collaboratively covers the grid and time point with the lowest quality as a grid-time combination, and this grid-time combination covers the weakest spatiotemporal bottleneck point.

[0039] Grid-time combination: ;in, This represents the grid and time point with the lowest cooperative coverage quality identified in a continuous target grid chain, i.e., the grid-time combination; This represents the grid cell with the lowest cooperative coverage quality identified in a continuous target grid chain; Indicates that in the corresponding Time slot; Represents the set of meshes in a continuous target mesh chain; Represents the first in a continuous target mesh chain One grid cell; This represents the total number of grid cells in a continuous target grid chain; Represents a grid In the time slot The quality of collaborative coverage; Methods for obtaining the optimal schedule include: A scheduling search space is established for each sensor. The working status of the sensor in each time slot of the prediction day and the corresponding sensing radius level are combined and encoded as candidate scheduling individuals. A local scheduling population is maintained for each sensor. The candidate scheduling individuals are subjected to evolutionary search by iteratively performing selection, crossover, mutation and elite retention operations. During the evolution process, the improvement of collaborative coverage quality at the spatiotemporal bottleneck point is used as the optimization objective. Under the premise of ensuring that the optimization objective is not reduced, the collaborative coverage quality of the continuous target grid chain is further optimized, thus forming a lexicographical dual-objective optimization mechanism to obtain the local optimal scheduling of a single sensor. Lexical order bi-objective optimization mechanism: ;in, Indicates sensor The optimal schedule represents the final schedule obtained after optimization. Represents candidate scheduling individuals, and represents sensors. The working status (working / sleeping) and radius level combination scheme for all time slots during the day; Indicates the scheduling's impact on time and space bottlenecks. The measure of collaborative coverage quality improvement should prioritize maximizing this goal to ensure that bottlenecks are covered as fully as possible. Indicates that without reducing Under the premise of scheduling A metric for improving the overall collaborative coverage quality of a continuous target mesh chain, used to optimize the coverage performance of the entire mesh chain; Maximize first ,exist Maximize under non-inferior conditions ; For candidate scheduling individuals whose energy consumption exceeds the available energy budget within the energy balance cycle during the evolution process, the candidate scheduling individuals with the lowest energy cost-effectiveness are gradually reduced or their work level is lowered until the available energy budget is met, based on the unit energy coverage gain corresponding to each time slot and each perception radius level, so as to achieve scheduling feasibility repair. Schedule feasibility fixes: ;in, This represents the benefit per unit of energy consumed, and the bottleneck coverage gain per unit of energy consumed. Indicates sensor In the time slot Using radius level Bottleneck The marginal coverage gain is the incremental contribution of the working unit to the collaborative coverage quality of the bottleneck point. Indicates sensor In the time slot Using radius level Energy consumption during operation, including operating energy consumption and the minimum energy required to maintain sensor operation; During the evolution of each sensor, elite scheduling summary information is exchanged at low frequency. When the evolution reaches the preset termination condition, the local optimal schedules obtained under different energy budget conditions are fused and scored to obtain the comprehensive score of the individual schedule. The score comprehensively considers the degree of coverage improvement of spatiotemporal bottlenecks, the overall collaborative coverage quality of continuous target grid chains, and energy utilization efficiency. Finally, the local optimal schedule with the highest comprehensive score of the individual schedule is selected as the optimal schedule of each sensor.

[0040] The individual scheduling score is: ;in, Indicates scheduling The combined target value under the two energy budget scenarios is used to measure the overall performance of the schedule under different energy conditions; Indicates the target index. , The corresponding bottleneck coverage improvement targets, The corresponding target is to improve the overall coverage quality of the continuous target grid chain; This indicates that under a conservative energy budget, the scheduling For the first The values ​​of each target are used to evaluate scheduling performance under conditions of low available energy. This indicates a conservative energy budget. Under the conditions, scheduling For the first The values ​​of each target are used to evaluate scheduling performance under conditions of low available energy. Indicates the median energy budget Under the conditions, scheduling For the first The values ​​of the targets are used to evaluate scheduling performance under moderate available energy conditions; This represents the fusion weight, used to adjust the conservative budget. Compared with median budget Final robustness score The proportion of influence; This represents a conservative estimate of the available energy level, while the other represents an estimate of the sensor's available energy under the most unfavorable conditions. This represents the median available energy level, indicating an estimate of the sensor's available energy under moderate conditions. It should be noted that during the scheduling optimization process, two quantifiable energy budget scenarios are pre-defined for each sensor: a conservative budget and a moderate budget. The conservative budget represents the minimum available energy of the sensor under the most unfavorable weather conditions during the day, which can be obtained by accumulating the low quantile value (e.g., the 10th percentile) of the intraday predicted power curve. The moderate budget represents the median available energy of the sensor under typical weather conditions during the day, which can be obtained by accumulating the median value (50th percentile) of the intraday predicted power curve. During the scheduling fusion process, the performance values ​​of the scheduling under each objective are first calculated according to the two budget scenarios. That is, under the conservative budget, the degree of improvement of bottleneck coverage and the overall collaborative coverage quality of the target chain are measured by the scheduling, and the corresponding target performance values ​​are calculated similarly under the moderate budget. Then, according to the fusion weight, the target performance values ​​under the conservative budget and the moderate budget are weighted and combined to obtain the robustness score of the scheduling under the two budgets, thereby quantifying the comprehensive performance of the scheduling under different energy availability conditions. Finally, based on the robustness score, the local optimal schedule with the highest score is selected as the final working schedule of the sensor to ensure that the coverage improvement of bottleneck points, the overall collaborative coverage quality of continuous target grid chains, and energy utilization efficiency can be taken into account under different energy conditions.

[0041] The methods for each sensor to perform monitoring tasks and report its operating status according to the optimal work schedule include: Each sensor performs monitoring tasks according to the optimal schedule obtained, within each time slot of the day, and according to the sensing radius level determined by the schedule. There are preset working time slots and sleeping time slots. Within the working time slots, the sensors collect monitoring data and detect targets in the grid cells covered by the continuous target grid chain. During the sleep time slot, the sensor enters standby mode. At the same time, the sensor reports the collected monitoring data and its own operating status to the network control terminal through a preset communication strategy. The monitoring data includes the collection time and detection results. The operating status includes the working status, sensing radius level and remaining energy.

[0042] It should be noted that in this invention, the daytime is divided into several discrete time slots, each with a fixed length. Based on the energy budget of each sensor, the intraday power prediction curve, and the coverage schedule, the system pre-determines the working time slot and the sleep time slot for each sensor within the day. The working time slot refers to the period during which the sensor needs to activate its sensing mode and perform monitoring tasks; the sleep time slot refers to the period during which the sensor does not need to perform monitoring and enters a low-power standby state. By pre-dividing the time slots, photovoltaic energy can be fully utilized and energy consumption can be saved while ensuring continuous target grid chain coverage, thereby extending the stable operating time of the sensor network.

[0043] During the working time slot, the sensors activate the corresponding sensing modes according to the sensing radius level determined by the schedule, performing data acquisition and target detection on the grid cells covered by the continuous target grid chain. The acquired data includes monitoring time and target detection results, which can be used for real-time coverage assessment and subsequent scheduling optimization.

[0044] During the sleep time slot, the sensor enters standby mode to reduce energy consumption. During this period, the sensor still needs to report the collected monitoring data and its own operating status to the network control terminal through a preset communication strategy. This communication strategy includes data aggregation, compression, and timed transmission mechanisms to ensure data upload is completed under conditions of minimal energy consumption. The sensor's reported operating status includes its current working status, sensing radius level, and remaining energy, enabling the network control terminal to perform overall status monitoring and scheduling adjustments.

[0045] The preset cosine similarity threshold is set by staff based on historical data analysis results. This historical analysis process includes the system collecting the cosine similarity of multiple sample vectors and calculating their average value as a reference to obtain the preset cosine similarity threshold. Similarly, the preset spatial distance threshold and the preset collaborative detectable potential threshold are also set by staff based on the system's historical operating data and specific application scenario requirements, and can be adjusted by staff during system operation according to actual conditions.

[0046] This embodiment introduces a tail risk metric to evaluate the photovoltaic power trajectory across multiple intraday scenarios. This quantifies power deviations under extremely unfavorable weather conditions, generates predictive power curves with controllable risks, and reduces the risk of power shortfalls and coverage interruptions. By comprehensively considering energy utilization efficiency and curtailment losses, it achieves efficient utilization of photovoltaic power generation resources and reduces energy waste. Tail risk perception scoring is applied to the photovoltaic power trajectories across multiple scenarios, and the generated predictive power curves take into account power fluctuations under different weather conditions, providing dynamic and quantifiable input for wireless sensor network coverage scheduling.

[0047] While ensuring sufficient sensor energy, the work schedule is dynamically adjusted through a risk-controlled energy budget and power prediction to balance coverage quality and energy efficiency, thereby improving the long-term operational performance of the network. Feasible sensing radius levels that meet the daily energy budget are selected, and the single node with the largest detection contribution is chosen as the effective contribution, ensuring optimal sensor coverage and avoiding energy waste or coverage interruptions. The collaborative detectable potential of grid cells with effective detection contributions from all sensors is accumulated, providing a quantitative basis for coverage scheduling and helping to identify weak areas and optimize scheduling. The collaborative detectable potential of the grid under different energy conditions is dynamically evaluated, allowing for flexible adjustment of sensor operating modes to balance coverage quality and energy consumption, improving the long-term stability of the network. The quantified results are used to construct continuous target grid chains, selecting the path with the maximum coverage potential under spatial continuity constraints, improving coverage effectiveness and task completion rate in the monitored area. Example 2

[0048] Please see Figure 2 As shown, parts not described in detail in this embodiment are described in Embodiment 1. A wireless sensor network coverage scheduling system for energy harvesting is provided, including: The risk perception and prediction module is used to construct intraday multi-scenario power trajectories based on historical meteorological and photovoltaic data collected by various sensors, introduce tail risk measurement to evaluate the trajectories, output intraday predicted power curves with controllable risks, and calculate the daily energy budget for each sensor. The energy balance construction module is used to divide the intraday time slot into different energy balance periods based on the intraday predicted power curve, and to establish available energy constraints within the balance period. The target chain selection module is used to grid the preset monitoring area, evaluate the collaborative detectability potential of the grid by combining the energy budget of each sensor, and select the grid that meets the continuity constraint in the preset spatial column structure to form a continuous target grid chain. The spatiotemporal bottleneck identification module is used to calculate the cooperative coverage quality of each grid in the target grid chain in each time slot, search for the grid-time combination with the smallest cooperative coverage quality within the range of the target grid chain and time slot combination, and identify the spatiotemporal bottleneck point with the weakest coverage. The distributed scheduling module is used to construct work schedules for each sensor under the constraint of available energy, with the collaborative coverage quality of spatiotemporal bottlenecks as the primary objective; it also repairs schedules that exceed energy limits, and performs fusion scoring on schedules under different energy budgets to obtain the optimal schedule; each sensor executes monitoring tasks according to the optimal work schedule and reports its operating status.

[0049] Since the electronic device described in this embodiment is the one used to implement the wireless sensor network coverage scheduling method and system based on energy harvesting in the embodiments of this application, those skilled in the art can understand the specific implementation and various variations of the electronic device in this embodiment based on the wireless sensor network coverage scheduling method and system based on energy harvesting in the embodiments of this application. Therefore, how the electronic device implements the method in the embodiments of this application will not be described in detail here. As long as those skilled in the art implement the electronic device used in the wireless sensor network coverage scheduling method and system based on energy harvesting in the embodiments of this application, it falls within the scope of protection of this application.

[0050] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0051] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A wireless sensor network coverage scheduling method for energy harvesting, characterized in that, include: S1. Based on the historical meteorological and photovoltaic data collected by each sensor, construct the intraday multi-scenario power trajectory, introduce tail risk measurement to evaluate the trajectory, output the intraday predicted power curve with controllable risk, and calculate the daily energy budget of each sensor. S2. Based on the intraday predicted power curve, the intraday time slot is divided into different energy balance cycles, and available energy constraints are established within the balance cycle. S3. Grid the preset monitoring area, evaluate the collaborative detectability potential of the grid by combining the energy budget of each sensor, and select the grid that meets the continuity constraint in the preset spatial column structure to form a continuous target grid chain. S4. Calculate the cooperative coverage quality of each grid in the target grid chain in each time slot, search for the grid-time combination with the smallest cooperative coverage quality within the range of the target grid chain and time slot combination, and identify the spatiotemporal bottleneck point with the weakest coverage. S5. Under the constraint of available energy, with the collaborative coverage quality of spatiotemporal bottlenecks as the primary objective, construct a work schedule for each sensor; repair the schedule that exceeds the energy limit, and perform fusion scoring on the schedule under different energy budgets to obtain the optimal schedule; each sensor executes monitoring tasks according to the optimal work schedule and reports its operating status.

2. The wireless sensor network coverage scheduling method for energy harvesting according to claim 1, characterized in that, The method for constructing intraday multi-scenario power trajectories includes: Continuous data collection is achieved through meteorological sensors and photovoltaic power output monitoring units deployed in the monitoring area. The meteorological sensors are used to acquire meteorological data, and the photovoltaic power output monitoring units are used to acquire photovoltaic power output data at the corresponding time. A unified time identifier is assigned to both the meteorological data and the photovoltaic power output data. The meteorological data and photovoltaic output power data are processed by timestamp alignment, outlier removal, missing value imputation and unified sampling interval to form a one-to-one corresponding sample sequence at the same time scale; and the meteorological data and photovoltaic output power data are processed by standard deviation normalization to construct a meteorological state index characterizing the intensity of the comprehensive meteorological effect, and the changing trend of the meteorological state index is extracted. The normalized meteorological state indicators and their changing trends are used as historical samples. The mean, variance, and quantiles of the historical samples in terms of indicator values ​​and rates of change are statistically analyzed. The indicator values ​​refer to the meteorological state indicators at a single moment, and the rates of change refer to the rates of change of the meteorological state indicators between consecutive time slots. By representing each historical sample as a sample vector, which includes the meteorological state index value of the sample in each time slot and the corresponding rate of change of the meteorological state index, the cosine similarity between any two sample vectors is calculated. If the cosine similarity of two sample vectors is greater than or equal to the preset cosine similarity threshold, the two samples are determined to have similar meteorological action mechanisms and photovoltaic power response characteristics. Based on the cosine similarity calculation results, all historical samples are divided into different meteorological scene types. Each meteorological scene type corresponds to a set of similar samples, and scene identifiers and frequency weights are assigned. For each meteorological field scenario type, combined with the theoretical irradiance law corresponding to the intraday solar altitude angle change, a theoretical power output envelope curve under the condition of no meteorological disturbance is constructed, and an intraday power trajectory is generated based on the power change characteristics of historical samples. The trajectories of each scenario are combined in chronological order and branched according to the frequency of occurrence to form a multi-scenario photovoltaic power scenario tree within a day. Each branch represents a possible power evolution path, thus forming a multi-scenario photovoltaic power trajectory within a day.

3. The wireless sensor network coverage scheduling method for energy harvesting according to claim 2, characterized in that, The method for calculating the daily energy budget for each sensor includes: For each candidate power trajectory in the multi-scenario photovoltaic power trajectory within the day, a tail risk metric is introduced to evaluate the candidate power trajectory, and a comprehensive evaluation function is formed by combining energy utilization utility and curtailment loss. The risk perception score of each candidate power trajectory is then obtained through the comprehensive evaluation function. Based on the risk perception score, a candidate power trajectory with the highest energy utilization efficiency, the lowest curtailment loss, and controllable tail risk is selected to generate an intraday predicted power curve. Based on the intraday predicted power curve, the power value corresponding to each sensor is obtained in each time slot, and combined with the rated capacity of each sensor, the daily energy budget of each sensor is calculated.

4. The wireless sensor network coverage scheduling method for energy harvesting according to claim 3, characterized in that, The method for establishing available energy constraints within the equilibrium period includes: The intraday predicted power curve is divided into different continuous time slots according to the time series, and each time slot corresponds to a fixed time interval. Based on the intraday predicted power curve, adjacent time slots are divided into different energy balance periods according to the cumulative available energy of the sensor, so that the total available energy in each balance period meets the sum of the predicted power demand in that energy balance period. Within each energy balance cycle, available energy constraints are established for each sensor. These constraints include that the total energy consumption of the sensor within the energy balance cycle is less than or equal to the available energy budget of the sensor within that energy balance cycle, and that the power consumption of the sensor within each time slot is less than or equal to the predicted power of the sensor corresponding to that time slot, thus forming a scheduling time organization framework.

5. The wireless sensor network coverage scheduling method for energy harvesting according to claim 4, characterized in that, The methods for assessing the collaborative detectability potential of the budget assessment grid include: The preset monitoring area is divided into different grid units according to the preset spatial resolution; based on the spatial positional relationship between the sensor and the grid unit and the preset probability perception model, the contribution of each sensor to the single-node detection of the grid unit at different preset perception radius levels is calculated. Obtain the daily energy budget and the corresponding working energy consumption of each sensing radius level for each sensor; for each grid cell, select feasible radius levels from the set of selectable sensing radius levels for each sensor that satisfy the requirement that the daily energy budget is not less than the energy consumption of that sensing radius level; Then, the radius level that makes the largest contribution to the detection of a single node in the grid cell is selected from the feasible radius levels as the effective detection contribution of the sensor to the grid cell; the effective detection contributions of all sensors are accumulated to obtain the cooperative detectable potential of the grid cell under energy constraints.

6. The wireless sensor network coverage scheduling method for energy harvesting according to claim 5, characterized in that, The method for forming a continuous target mesh chain includes: For each grid cell, grid cells that meet the energy feasibility criteria are selected as candidate grid cells based on the cooperative detectability potential of each grid cell under energy constraints and the minimum energy required for each sensor to maintain operation on each grid cell. One target grid is selected sequentially from the candidate grid cells in each column. During the selection process, the spatial distance between the selected target grids in adjacent columns is less than or equal to a preset spatial distance threshold. At the same time, grids with cooperative detectability potential greater than the preset cooperative detectability potential threshold are selected first, thereby obtaining a continuous grid sequence from the first column to the last column, which forms a continuous target grid chain.

7. The wireless sensor network coverage scheduling method for energy harvesting according to claim 6, characterized in that, The method for identifying the spatiotemporal bottleneck point with the weakest coverage includes: For each grid in the formed continuous target grid chain, the single-node detection contributions of each sensor are fused according to the preset collaborative fusion rules in each time slot of the prediction day to obtain the collaborative coverage quality of the grid in that time slot. Within the range of continuous target grid chains and time slots, the search collaboratively covers the grid and time point with the lowest quality as a grid-time combination, and this grid-time combination covers the weakest spatiotemporal bottleneck point.

8. The wireless sensor network coverage scheduling method for energy harvesting according to claim 7, characterized in that, The method for obtaining the optimal schedule includes: A scheduling search space is established for each sensor. The working status of the sensor in each time slot of the prediction day and the corresponding sensing radius level are combined and encoded as candidate scheduling individuals. A local scheduling population is maintained for each sensor. The candidate scheduling individuals are subjected to evolutionary search by iteratively performing selection, crossover, mutation and elite retention operations. During the evolution process, the improvement of collaborative coverage quality at the spatiotemporal bottleneck point is used as the optimization objective. Under the premise of ensuring that the optimization objective is not reduced, the collaborative coverage quality of the continuous target grid chain is further optimized, thus forming a lexicographical dual-objective optimization mechanism to obtain the local optimal scheduling of a single sensor. For candidate scheduling individuals whose energy consumption exceeds the available energy budget within the energy balance cycle during the evolution process, the candidate scheduling individuals with the lowest energy cost-effectiveness are gradually reduced or their work level is lowered until the available energy budget is met, based on the unit energy coverage gain corresponding to each time slot and each perception radius level, so as to achieve scheduling feasibility repair. During the evolution of each sensor, elite scheduling summary information is exchanged at low frequency. When the evolution reaches the preset termination condition, the local optimal schedules obtained under different energy budget conditions are fused and scored to obtain the comprehensive score of the individual schedule. Finally, the local optimal schedule with the highest comprehensive score of the individual schedule is selected as the optimal schedule of each sensor.

9. The wireless sensor network coverage scheduling method for energy harvesting according to claim 8, characterized in that, The method for each sensor to perform monitoring tasks and report its operating status according to the optimal work schedule includes: Each sensor performs monitoring tasks according to the optimal schedule obtained, within each time slot of the day, based on the sensing radius level determined by the schedule. There are preset working time slots and sleeping time slots. During the working time slot, the sensor collects monitoring data and detects targets in the grid cells covered by the continuous target grid chain. During the sleeping time slot, the sensor enters standby mode, and at the same time, the sensor reports the collected monitoring data and its own operating status to the network control terminal through a preset communication strategy.

10. A wireless sensor network coverage scheduling system for energy harvesting, used to implement the wireless sensor network coverage scheduling method for energy harvesting as described in any one of claims 1 to 9, characterized in that, include: The risk perception and prediction module is used to construct intraday multi-scenario power trajectories based on historical meteorological and photovoltaic data collected by various sensors, introduce tail risk measurement to evaluate the trajectories, output intraday predicted power curves with controllable risks, and calculate the daily energy budget for each sensor. The energy balance construction module is used to divide the intraday time slot into different energy balance periods based on the intraday predicted power curve, and to establish available energy constraints within the balance period. The target chain selection module is used to grid the preset monitoring area, evaluate the collaborative detectability potential of the grid by combining the energy budget of each sensor, and select the grid that meets the continuity constraint in the preset spatial column structure to form a continuous target grid chain. The spatiotemporal bottleneck identification module is used to calculate the cooperative coverage quality of each grid in the target grid chain in each time slot, search for the grid-time combination with the smallest cooperative coverage quality within the range of the target grid chain and time slot combination, and identify the spatiotemporal bottleneck point with the weakest coverage. The distributed scheduling module is used to construct work schedules for each sensor under the constraint of available energy, with the collaborative coverage quality of spatiotemporal bottlenecks as the primary objective; it also repairs schedules that exceed energy limits, and performs fusion scoring on schedules under different energy budgets to obtain the optimal schedule; each sensor executes monitoring tasks according to the optimal work schedule and reports its operating status.

Images