Sea-land-air-diving integrated inspection scheduling method

Abstract

The invention discloses a sea-land-air-diving integrated inspection scheduling method, relates to the technical field of electrical equipment inspection, and solves the problem that the prior art is lack of collaborative assessment of multi-dimensional dynamic risks and cross-domain integrated scheduling optimization capability. Through fusion of multi-source heterogeneous data, risk quantitative prediction transition from single experience judgment to data driving is realized, and composite risk levels of sea-land-air-dive domains in future time periods can be actively identified and evaluated instead of passive response. In addition, an economic, time, energy consumption and risk avoidance four-dimensional weighting objective function is also established, a unified collaborative scheduling language and optimization criterion are provided for heterogeneous inspection equipment, and an unmanned aerial vehicle, an unmanned ship, an underwater robot and the like are converted into a global linkage organic whole from an isolated operation unit. The model not only can dynamically adjust the inspection strategy according to the real-time environment change to realize the balance between the safety risk and the resource efficiency, but also can improve the prediction precision by continuously iterating historical data.

Description

Technical Field

[0001] This invention relates to the field of power equipment inspection technology, and in particular to an integrated sea, land, air and underwater inspection and scheduling method. Background Technology

[0002] With the accelerated global energy transition, offshore wind power, as an important carrier of clean energy, is rapidly developing towards deep-sea and large-scale operations, facing unprecedented complexity and challenges in its operation and maintenance management. Offshore wind farms typically consist of dozens to hundreds of wind turbines, distributed over hundreds of square kilometers of sea area. Inspection tasks must simultaneously cover multiple dimensions, including the offshore wind turbines themselves, submarine cables, substation platforms, and onshore control centers. These devices operate in harsh marine environments characterized by high salt spray, strong corrosion, and the combined effects of wind, waves, and currents, resulting in a significantly higher failure rate than onshore facilities. Furthermore, frequent extreme weather events such as typhoons, heavy fog, and thunderstorms pose severe challenges to traditional, single-mode inspection methods. Currently, while intelligent equipment such as drones, unmanned vessels, and underwater robots are gradually being applied to wind farm inspections, the fragmented operation mode leads to dispersed resource allocation, low collaborative efficiency, and difficulty in forming a unified situational awareness and decision-making system. Against this backdrop, there is an urgent need to build an intelligent scheduling method that can integrate multi-source heterogeneous data (including historical meteorological data, equipment fault records, flight logs, power consumption data and task assignment records) and coordinate multi-dimensional spatial inspection resources across sea, land, air and submarine. By accurately predicting the dynamic risks in different regions, it can achieve collaborative optimization and integrated scheduling of heterogeneous equipment, thereby improving the safety, economy and timeliness of offshore wind farm operation and maintenance.

[0003] The core deficiency of existing technologies lies in the lack of collaborative assessment and cross-domain integrated scheduling optimization capabilities for multi-dimensional dynamic risks. Traditional inspection and scheduling methods rely heavily on manual experience or static decision-making based on local data from a single equipment type, failing to effectively integrate multi-source heterogeneous information such as weather forecasts, equipment health status, and historical failure modes. This results in one-sided risk identification and insufficient assessment accuracy. Particularly in offshore wind power scenarios, existing solutions still treat drones, unmanned vessels, and underwater robots as independent operating units, lacking cross-domain collaborative mechanisms and unified risk quantification standards, and are unable to dynamically adjust inspection strategies according to real-time environmental changes. This fragmented scheduling model makes the system slow to respond to sudden severe weather or equipment failures, often falling into the dilemma of "overly aggressive operations leading to safety accidents" or "overly conservative practices resulting in idle resources."

[0004] Therefore, an integrated sea, land, air, and underwater inspection and dispatching method is needed. Summary of the Invention

[0005] To address the lack of collaborative assessment and cross-domain integrated scheduling optimization capabilities for multi-dimensional dynamic risks in existing technologies, this invention provides an integrated sea, land, air, and submarine inspection scheduling method. This method constructs a comprehensive weighted cost objective function, incorporating four major cost elements—economic, time, energy consumption, and risk avoidance—into a unified optimization framework. Furthermore, it introduces a historical data-driven risk prediction model, achieving coordinated scheduling and dynamic optimization of inspection resources across all sea, land, air, and submarine domains. This precisely solves the fundamental shortcomings of traditional methods, such as poor adaptability and low collaborative efficiency in complex environments. The specific technical solution is as follows: A method for integrated sea, land, air, and underwater inspection and dispatch includes the following steps: Collect and organize historical meteorological data, equipment failure records, UAV flight logs, power consumption data, and task assignment records; Assess the potential impact of severe weather and equipment failure risks that different regions may face when carrying out future missions, calculate the comprehensive risk value for all mission areas, and classify the risk levels. A scheduling optimization model is constructed with the goal of minimizing the overall weighted cost. The risk prediction results are input into the scheduling optimization model to solve and generate the optimal UAV scheduling scheme. The expression of the objective function is as follows: In the formula, It is the total weighted average cost; Is the type as A collection of drones, These are the weighting coefficients for economic cost, time cost, energy cost, and risk aversion cost, respectively, and they are added together to equal 1. For economic costs; For time cost; For energy consumption costs, To mitigate the costs of risk.

[0006] The preferred risk level classification process is as follows: Obtain severe weather risk assessment indicators, including at least the frequency of wind speed exceeding limits, the frequency of rainfall exceeding limits, the frequency of visibility falling below the threshold, and the frequency of lightning activity; Obtain equipment failure risk assessment indicators, including at least equipment failure rate, mean time between failures, and mean time to repair. The entropy value and weight of each risk indicator are calculated, and then the overall risk value is calculated, as follows: In the formula, Indicates the first i The overall risk value of each region; Indicates the first j The weight of each indicator; Indicates the firsti The region in the first j Standardized values ​​for each risk indicator; The calculated comprehensive risk value They are divided into different risk levels.

[0007] Preferably, the scheduling optimization model also includes constraints, which include at least the following:

[0008] Preferably, the scheduling optimization model also includes constraints, including at least the task uniqueness constraint: each task area must be executed by one drone only once.

[0009] Preferably, the scheduling optimization model also includes constraints, including at least a drone type matching constraint: task allocation must take into account the suitability of the drone type. For example, underwater inspection tasks can only be assigned to submersible drones.

[0010] Preferably, the scheduling optimization model also includes constraints, including at least the drone's endurance constraint: the total energy consumption of the drone during the mission cannot exceed its battery capacity.

[0011] Preferably, the method further includes the following steps: based on historical data and future task predictions, running a charging station site selection optimization model to determine the optimal location and number of charging stations, wherein the charging station site selection optimization model is as follows: In the formula, This represents the comprehensive benefit index, which is the ultimate goal of model optimization. This indicates the task coverage rate, which is the proportion of the number of task areas served by the charging station to the total number of task areas. This represents the total cost, including the construction cost and long-term operating cost of the charging station; The risk cost is mainly considered in terms of the risk of mission failure due to unreasonable charging station layout and the risk of the drone running out of power. These are the weighting coefficients for task coverage, total cost, and risk cost, respectively.

[0012] Preferably, the charging station location optimization model also includes constraints, including at least a task coverage constraint: each task area must be covered by at least one charging station.

[0013] A computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device where the computer-readable storage medium is located to perform the integrated sea, land, air, and underwater inspection and scheduling method as described above.

[0014] A processor for running a program, wherein the program executes the integrated sea, land, air, and underwater inspection and scheduling method as described above.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention integrates heterogeneous data from multiple sources, including historical meteorological data, equipment failure data, flight logs, energy consumption data, and mission records. The solution achieves a leap from relying solely on experience-based judgment to data-driven quantitative risk prediction, proactively identifying and assessing the combined risk levels of various domains—sea, land, air, and underwater—in future timeframes, rather than passively responding. It innovatively establishes a four-dimensional weighted objective function encompassing economy, time, energy consumption, and risk avoidance, providing a unified collaborative scheduling language and optimization criteria for heterogeneous inspection equipment. This transforms drones, unmanned surface vessels, and underwater robots from isolated operational units into a fully interconnected organic whole. The model not only dynamically adjusts inspection strategies based on real-time environmental changes to achieve a balance between safety risks and resource efficiency but also continuously improves prediction accuracy through iterative historical data iteration. It fundamentally solves the inherent shortcomings of traditional technologies—the lack of cross-domain collaborative mechanisms and difficulty in adapting to complex dynamic environments—promoting inspection management in scenarios such as offshore wind farms from a fragmented approach to a new level of intelligent integration. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0017] Figure 1 This is a flowchart of the method of 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, not all, of the embodiments of the present invention. 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.

[0019] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0020] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0021] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0022] Example 1 In one embodiment of the present invention, an integrated sea, land, air, and underwater inspection and scheduling method is provided, aiming to construct a scientific and objective risk prediction and quantitative assessment model. The core task of risk prediction is to assess the potential impact of severe weather and equipment failure risks that different regions may face when performing future tasks. By introducing the entropy weight method, this scheme can objectively assign weights to multi-dimensional risk indicators and ultimately calculate a comprehensive quantitative risk value. This value not only includes the probability information of the risk but also reflects the severity of the risk through a level classification, thereby achieving a comprehensive and accurate characterization of the risk.

[0023] The implementation steps of this embodiment are described in detail below: Step 1: Construct a risk indicator system To achieve accurate prediction and quantitative assessment of risks in UAV mission areas, a scientific and comprehensive risk indicator system is first needed. This system will primarily revolve around two core risk sources: the impact of severe weather and equipment failure risk. Each primary indicator will be further subdivided into multiple quantifiable secondary indicators to ensure the objectivity and accuracy of the assessment. This hierarchical indicator system not only facilitates a deeper understanding of the composition of risks but also provides a solid data foundation for subsequent use of objective weighting models such as the entropy weight method. Specifically: S101: Obtain severe weather risk assessment indicators; Severe weather is the primary external factor affecting the flight safety of unmanned aerial vehicles (UAVs), and its complexity and variability place high demands on risk assessment. To quantify the risks posed by severe weather, an assessment model incorporating multiple meteorological parameters is constructed. These meteorological parameters are selected based on their direct impact on UAV flight safety, as follows: 1. Frequency of Wind Speed ​​Exceeding Limits. Wind speed is a crucial indicator. Excessive wind speed not only increases the energy consumption of drones but may also lead to loss of control. Therefore, a wind speed threshold (e.g., 10 m / s) is set, and the frequency of wind speed exceeding this threshold in the area within a specific time period is counted as the frequency of wind speed exceeding limits.

[0024] 2. Frequency of Rainfall Exceeding Limits. Rainfall is also a critical parameter; heavy rain can affect the performance of drone sensors and even damage the equipment. Set a rainfall threshold (e.g., 5 mm / h) and calculate the frequency of rainfall exceeding the limit.

[0025] 3. Frequency of visibility below the threshold. Visibility is directly related to the visual perception ability of the drone operator or autonomous system. Flying under low visibility conditions is extremely risky. Therefore, the frequency of visibility below the threshold (such as visibility below 1 kilometer) is also a core evaluation indicator.

[0026] 4. Lightning activity frequency. Lightning activity poses a direct threat to the electronic equipment of drones. Statistics on lightning activity frequency can reflect the lightning risk in a region during a specific time period.

[0027] By comprehensively analyzing these secondary indicators, a comprehensive severe weather risk assessment model can be constructed, providing basic data for subsequent comprehensive risk calculations.

[0028] S102: Obtain equipment failure risk assessment indicators; Equipment failure is another major risk in drone operation. Its probability of occurrence is closely related to the equipment's health condition, usage frequency, and maintenance level, as detailed below: 1. Equipment Failure Rate. The equipment failure rate is the most intuitive indicator, and its calculation formula is: number of failures divided by the total number of equipment operations, then multiplied by 100%. This indicator directly reflects the frequency of equipment failures during operation.

[0029] 2. Mean Time Between Failures (MTBF). The MTBF is calculated as: total uptime divided by the number of failures. A higher MTBF indicates a longer average uptime between two failures, and thus higher reliability.

[0030] 3. Mean Time To Repair (MTTR). The formula is: Total time to repair faults divided by the number of faults. MTTR reflects the efficiency of equipment recovery after a fault. A shorter MTTR indicates stronger maintenance response and repair capabilities, and less impact on task continuity.

[0031] 4. Performance metrics related to network devices, such as CPU and memory utilization, packet loss rate, latency, and jitter, can reflect the health status of the drone in terms of data links and computing load, thereby providing a more comprehensive assessment of its potential failure risks.

[0032] Step 2: Calculate the comprehensive risk value based on the entropy weight method; To achieve objective risk quantification, this scheme employs the Entropy Weight Method to calculate the comprehensive risk value. The Entropy Weight Method is an objective weighting method based on information theory, which determines weights by calculating the dispersion of indicator data. The greater the dispersion of an indicator, the higher its variability and uncertainty within the assessment system, and the more information it provides; therefore, its weight should be correspondingly greater. This method avoids the potential personal bias and arbitrariness inherent in subjective weighting methods (such as expert scoring), making the risk assessment results more scientific and reliable. This embodiment uses severe weather risk and equipment failure risk as two core assessment indicators, determining their objective weights in the comprehensive risk value using the Entropy Weight Method, thereby obtaining a quantitative result that reflects both the probability and severity of risk. Specifically: S201: Data standardization and forward processing; Before applying the entropy weight method, the original risk data must be preprocessed to eliminate the impact of differences in the dimensions and magnitudes of different indicators and to ensure that all indicators have a consistent influence on the final assessment result. Since the risk indicators in this embodiment (severe weather risk, equipment failure risk) are all negative indicators (i.e., the larger the value, the higher the risk), positive transformation is required. The purpose of positive transformation is to convert negative indicators into positive indicators, so that the larger the converted value, the lower the risk (or the higher the safety), thus maintaining consistency with the calculation logic of the entropy weight method. Specifically, the Min-Max Normalization method is used to standardize and positively transform the data. For negative indicators, the standardization and positive transformation formulas are: In the formula, Indicates the first i The region in the first j The standardized value of each risk indicator ranges from [0, 1]. The larger the value, the lower the risk of that indicator in that region. Indicates the first i The region in the first j The raw data values ​​for each risk indicator; Indicates the first in all regions j The maximum value of each risk indicator; Indicates the first in all regions j The minimum value of each risk indicator.

[0033] This formula linearly maps the original risk data to the [0, 1] interval, achieving a positive transformation. For example, the higher the original risk value of a region, the higher its standardized risk value will be. The smaller the value, the better, and vice versa. This step forms the basis for subsequent entropy and weight calculations, ensuring the scientific rigor and consistency of data processing.

[0034] S202: Calculate the entropy value and weight; After completing data standardization and positive conversion, the next step is to calculate the entropy value and weight of each risk indicator. The entropy value reflects the dispersion of the indicator data, while the weight quantifies the importance of that indicator within the overall risk assessment system. The calculation process is as follows: 1. Calculate the first... j The first indicator i The characteristic proportion of each region : In the formula, Indicates the first j The first indicator i The characteristic proportions of each region; Indicates the total number of regions; Indicates the first i The region in the first j Standardized values ​​for each risk indicator.

[0035] To avoid the mathematical error of log(0) in subsequent logarithm calculations, this embodiment also... The correction is made by replacing its 0 value with a very small positive number (e.g., 10). -8 ).

[0036] 2. Calculate the first... j The entropy value of each indicator: Entropy measures the degree of disorder or uncertainty in indicator data. A higher entropy value indicates more concentrated data, less variability, and less information provided, as detailed below: In the formula, Indicates the first j The entropy value of each indicator, is the Boltzmann constant, with values ​​ranging from [0,1].

[0037] 3. Calculate the first... j Coefficient of difference of each indicator : The coefficient of difference reflects the effective information content and importance of an indicator. A larger coefficient of difference indicates a greater information content of the indicator, and therefore a higher weight should be assigned, as detailed below: 4. Calculate the first... j Weight of each indicator : The weights are the normalized results of the difference coefficients, representing the relative importance of the indicator in the overall evaluation, as detailed below: In the formula, This represents the total number of risk indicators. In this embodiment, n=2, which represents the risk of severe weather and the risk of equipment failure.

[0038] Through the above steps, the objective weight of each risk indicator is obtained. These weights are entirely determined by the dispersion of the data itself and can truly reflect the contribution of different risk factors to the overall risk level.

[0039] S203: Comprehensive Risk Value Calculation and Risk Level Classification; After obtaining the weights of each risk indicator, the overall risk value for each region can be calculated. This value is the final quantitative indicator for measuring the overall risk level of the region.

[0040] The formula for calculating the overall risk value is as follows: In the formula, Indicates the first i The overall risk value of each region; Indicates the first j The weight of each indicator; Indicates the first i The region in the first j Standardized values ​​for each risk indicator; It is worth noting that, due to It has undergone positive transformation; the higher the value, the lower the risk. Therefore, the calculated comprehensive risk value... The larger the value, the lower the overall risk level and the higher the safety of the area. This is contrary to our usual understanding of "risk value" (the higher the value, the higher the risk). To align with conventional understanding, we can... Perform another transformation, for example, define a new composite risk value. ,so The higher the value, the higher the risk.

[0041] The risk levels are classified as follows: To more intuitively understand and apply the risk assessment results, it is necessary to use the calculated comprehensive risk value. The risk levels are divided into different categories. These categories can be categorized using various methods, such as interval division, quantile division, or custom threshold division based on business needs.

[0042] For example, risks can be divided into four levels: Low Risk: 0≤ <0.25; Medium Risk: 0.25 ≤ <0.5; High Risk: 0.5≤ <0.75; Extremely High Risk: 0.75≤ ≤1.0; This tiered classification enables decision-makers to quickly identify high-risk areas that require focused attention, and provides clear priority guidance for subsequent drone scheduling and charging station deployment.

[0043] Step 3: Optimize the scheduling of UAVs across sea, land, air, and submarine based on risk prediction, as detailed below: S301: Modeling the UAV scheduling problem; To achieve efficient scheduling of four types of UAVs—sea, land, air, and underwater—this embodiment constructs a multi-objective, multi-constraint mathematical optimization model. The core objective of this model is to minimize the comprehensive weighted cost, comprised of economic cost, time cost, energy cost, and risk avoidance cost, while meeting mission requirements and various constraints. The model's inputs include the risk level of each region obtained from the risk prediction module, the geographical type of the mission area (sea, land, air, underwater), the performance parameters of various UAVs (such as endurance, speed, payload, and operational cost), and the number of available UAVs. By establishing an accurate mathematical model, the complex scheduling problem can be transformed into a solvable optimization problem, laying the foundation for subsequent use of intelligent optimization algorithms (such as genetic algorithms) to find the optimal scheduling scheme.

[0044] First, define the decision variables as follows: 1. Task allocation variables : This is a three-dimensional binary variable used to indicate whether the first... i The inspection task for the area was assigned to the first... k A drone, and the drone belongs to type [missing information]. t The details are as follows: In the formula, i ∈{1, 2, ..., N} represents the index of the task region. N This represents the total number of task regions. k ∈{1, 2,..., M} represents the index of the drone. M This represents the total number of available drones; t ∈{sea, land, air, underwater} indicates the type of drone.

[0045] 2. Path variables: This is a three-dimensional binary variable used to represent a drone. k From the region i Fly directly to the area j To execute a task, the following is indicated: This variable is mainly used to construct the drone's flight path and to calculate flight time and energy consumption in the objective function.

[0046] 3. Time variable T ik : T ik Indicates drone k Arrival Area i And the start time of task execution. This is a continuous variable used to handle time window constraints and calculate task completion time.

[0047] By combining these decision variables, a complete drone scheduling scheme can be described, including which drone will perform each task, the drone's flight path, and the time sequence of task execution.

[0048] S302: Construct an objective function with the goal of minimizing the overall weighted cost.

[0049] The overall cost consists of four parts: economic cost, time cost, energy cost, and risk aversion cost. The objective function is the weighted sum of these four costs; by adjusting the weighting coefficients, the relative importance of different costs in the decision-making process can be reflected. The mathematical expression of the objective function is as follows: In the formula, It is the total weighted average cost; Is the type as A collection of drones, These are the weighting coefficients for economic cost, time cost, energy cost, and risk aversion cost, respectively, and they are added together to equal 1. For economic costs; For time cost; For energy consumption costs, To mitigate the costs of risk. The specific calculation methods for each cost are as follows: 1. Economic Costs : refers to drones k Direct costs incurred in performing the task include equipment depreciation, maintenance costs, and operator salaries; 2. Time cost This refers to the total time required to complete the mission, including flight time and operational time at the mission site.

[0050] 3. Energy consumption cost : refers to drones k Total electricity consumed during flight and operation.

[0051] 4. Risk avoidance costs Losses may result from drones performing missions in specific areas (especially high-risk areas).

[0052] By minimizing this comprehensive weighted cost objective function, a drone scheduling scheme that achieves the best balance between economic benefits, efficiency, energy consumption, and safety can be found.

[0053] S303: Set constraints, as follows: 1. Mission uniqueness constraint: Each mission area must be executed by one drone only once.

[0054] 2. Drone type matching constraint: Task assignment must take into account the suitability of the drone type. For example, underwater inspection tasks can only be assigned to submersible drones.

[0055] 3. Priority Constraints for High-Risk Areas: For areas designated as high-risk, specific types of drones (such as models with superior performance and stronger risk resistance) must be prioritized for inspection.

[0056] 4. Drone endurance constraints: The total energy consumption of a drone during mission execution must not exceed its battery capacity.

[0057] 5. Path flow balance constraint: Ensure that the flight path of the UAV is continuous, that is, the UAV must leave the area after entering the area (except for the starting point and the ending point).

[0058] 6. Time window constraint: Some tasks may need to be completed within a specific time window.

[0059] 7. Drone Number Constraint: The number of drones performing a mission simultaneously cannot exceed the total number of available drones.

[0060] By minimizing this comprehensive weighted cost objective function, a drone scheduling scheme that achieves the best balance between economic benefits, efficiency, energy consumption, and safety can be found.

[0061] S304: Set a drone assignment strategy based on risk level, as follows: Prioritizing specific types of drones in high-risk areas: Risk avoidance is a crucial objective in drone dispatching. To effectively address the challenges posed by high-risk areas, a drone assignment strategy based on risk levels is necessary. Specifically, mission areas can be categorized into high, medium, and low risk levels based on risk prediction results. For high-risk areas, drones with stronger resilience should be prioritized, such as those capable of flying in adverse weather conditions, possessing higher protection levels, or enhanced communication capabilities. This strategy can significantly improve mission success rates and reduce the probability of drone malfunctions or accidents during missions. For example, in high-wave maritime areas, "sea-type" drones with waterproof and corrosion-resistant capabilities can be prioritized; in urban areas with severe electromagnetic interference, "air-type" drones with interference-resistant communication links can be deployed. Through this targeted assignment, the safety of the drone system can be maximized while ensuring mission completion quality.

[0062] Cost-Effective Drone Selection: While prioritizing the deployment of specific drone types, cost-effectiveness must also be considered. Different types of drones differ in procurement, maintenance, and energy costs. Therefore, a trade-off must be struck between risk mitigation and cost control when assigning drones. For example, while high-performance drones are better suited for high-risk environments, their operating costs are also relatively higher. Therefore, under the premise of controllable risk, the most cost-effective drone type should be selected. To achieve this goal, a cost-effectiveness assessment model can be established that comprehensively considers drone performance, cost, and mission requirements to recommend the optimal drone type for each mission. For example, for low-risk, short-distance missions, a lower-cost "land" drone can be selected; while for high-risk, long-distance missions, a more powerful but also more expensive "air" drone can be chosen. Through this refined management, overall operating costs can be minimized while ensuring mission safety and efficiency.

[0063] S305: Solving for the optimal scheduling scheme This embodiment uses a genetic algorithm (GA) to find an approximate optimal solution to the problem. The specific steps are as follows: 1. Encoding: Representing the scheduling scheme in chromosome form, such as integer encoding.

[0064] 2. Initial Population: Randomly generate an initial scheduling scheme.

[0065] 3. Fitness Evaluation: Evaluate the merits of each solution based on the objective function (comprehensive weighted cost).

[0066] 4. Selection: Selecting parent individuals based on fitness, such as roulette wheel selection.

[0067] 5. Crossover: Exchanging parts of the chromosomes of a parent individual to produce offspring.

[0068] 6. Mutation: Randomly altering genes on chromosomes with a small probability to increase population diversity.

[0069] 7. Termination Condition: Repeat the iterations until the maximum number of iterations or the convergence condition is met.

[0070] Through the above iterative process, the genetic algorithm can gradually optimize the population and eventually converge to one or a set of high-quality scheduling schemes.

[0071] Example 2 This embodiment, based on Embodiment 1, comprehensively considers multiple constraints such as candidate deployment locations, service radius limitations, and budget limits, taking into account historical UAV power consumption and assignment data. Its core objective is to effectively balance construction costs and long-term operating expenses while achieving high coverage of the mission area, ultimately forming a site selection and quantity design scheme with the highest overall efficiency. The steps include: Step 4: Modeling the charging station site selection problem, as follows: S401: Construct decision variables, specifically including charging station site selection decision variables and task area service decision variables.

[0072] Let the set of candidate positions be The decision variable for charging station site selection is specifically represented as: whether a candidate location is selected as the construction site for a charging station. Defined as: =1 indicates that at the candidate position j Build charging stations; =0 indicates that it is not in the candidate position. j Build charging stations; The task area service decision variable represents whether a certain task area is served by a certain charging station. Let the set of task areas be... ,but Defined as: =1 indicates the task area i By the candidate position j The charging stations provide services; =0 indicates the task area i Not located in the candidate position jThe charging stations provide services.

[0073] By using these two sets of decision variables, the charging station site selection problem can be transformed into a typical integer programming problem, where the decision variables can only take the values ​​0 or 1. This allows the model to accurately describe the discrete decisions of "build" or "not build" and "serve" or "not serve".

[0074] S402: Construct an objective function with the goal of maximizing the comprehensive benefit index.

[0075] The core objective of drone charging station site selection is to maximize overall benefits while meeting all mission requirements. This overall benefit is the result of multi-objective optimization, requiring a balance between factors such as mission coverage, construction costs, operating costs, and risk mitigation. Therefore, the design of the objective function needs to quantify and weight these indicators with different dimensions to form a unified evaluation standard. The expression of the objective function is as follows: In the formula, This represents the comprehensive benefit index, which is the ultimate goal of model optimization. This indicates the task coverage rate, which is the proportion of the number of task areas served by the charging station to the total number of task areas. This represents the total cost, including the construction cost and long-term operating cost of the charging station; The risk cost is mainly considered in terms of the risk of mission failure due to unreasonable charging station layout and the risk of the drone running out of power. These are the weighting coefficients for task coverage, total cost, and risk cost, respectively.

[0076] By adjusting the weighting coefficients, the objective function can be flexibly adapted to different application scenarios and decision-making needs, thereby finding an optimal balance between coverage, cost, and risk.

[0077] Set the constraints as follows: 1. Task coverage constraint: Each task area must be covered by at least one charging station.

[0078] 2. Service radius constraint: Due to their limited endurance, drones can only perform tasks and recharge within a certain range. Therefore, the service range of charging stations is limited.

[0079] 3. Budget constraints: The construction and operation of charging stations require a large investment of funds, so the budget is an important factor that must be considered in the site selection plan.

[0080] 4. Charging station capacity constraints: Each charging station has a limited charging capacity and cannot provide charging services for drones indefinitely. By setting these constraints, the model can generate charging station site selection schemes that are both realistic and feasible.

[0081] Step 5: Predict power consumption based on historical data.

[0082] Accurate power consumption prediction is fundamental to optimizing charging station location. By analyzing historical drone dispatch and power consumption data, we can understand the energy consumption patterns in different mission areas, drone types, and mission modes, thus enabling scientific predictions of future power demand. This data-driven prediction method is more accurate than traditional experience-based estimations and can effectively avoid the problem of unreasonable charging station layout due to prediction errors.

[0083] This embodiment uses a time series model (such as ARIMA) to generate a predicted power consumption demand for a future period for each task area. These predicted values ​​will serve as important input parameters for the charging station site selection model, used to determine the capacity and layout of charging stations.

[0084] Step 6: Solve the objective function to obtain the optimal charging station location strategy.

[0085] For example, heuristic algorithms or metaheuristic algorithms are used to find approximate optimal solutions.

[0086] For example, it includes the following steps: S601: First, these candidate locations are screened, eliminating those that clearly do not meet the requirements. Candidate point screening based on coverage area is a commonly used preprocessing method. Its core idea is to exclude candidate points that cannot cover any mission area based on the UAV's maximum service radius. The specific steps are as follows: 1. Calculate the distance matrix: Calculate the distance between all candidate locations and all task regions.

[0087] 2. Set coverage radius: Based on the drone's endurance, set a maximum service radius R.

[0088] 3. Filter candidate points: For each candidate location, check if at least one task region is within its service radius. If a candidate location is more than R away from all task regions, then the candidate point is discarded.

[0089] Example 3 This embodiment is based on Embodiments 1 and 2, such as Figure 1As shown, the data-driven and dynamic linkage between the three modules—risk prediction, drone scheduling, and charging station site selection—forms a closed-loop decision optimization system. The output of each module serves as the key input for the next module, thereby ensuring that the overall system's decisions are coherent, coordinated, and optimal.

[0090] 1. Data Preparation and Model Initialization: Collect and organize historical meteorological data, equipment failure records, UAV flight logs, power consumption data, and task assignment records. Initialize the parameters in the risk prediction, scheduling optimization, and site selection optimization models.

[0091] 2. Risk Prediction and Assessment: Run the risk prediction module to calculate the comprehensive risk value of all task areas and classify the risk levels.

[0092] 3. Scheduling scheme generation and execution: Input the risk prediction results into the scheduling optimization model, use solvers such as genetic algorithms to generate the optimal UAV scheduling scheme, and issue execution instructions.

[0093] 4. Charging station site selection planning: Based on historical data and future task predictions, run the charging station site selection optimization model to determine the best location and number of charging stations, and carry out construction and deployment.

[0094] 5. System Operation and Data Feedback: After the system is put into operation, it continuously collects new task execution data, power consumption data, and risk event data to form a new historical dataset.

[0095] In summary, this technical solution, by integrating multi-source heterogeneous data such as historical meteorological data, equipment failure data, flight logs, energy consumption data, and mission records, achieves a leap from single-experience judgment to data-driven quantitative risk prediction. It proactively identifies and assesses the composite risk levels of various domains—sea, land, air, and underwater—in future timeframes, rather than passively responding. Its innovative four-dimensional weighted objective function—comprising economy, time, energy consumption, and risk avoidance—provides a unified collaborative scheduling language and optimization criteria for heterogeneous inspection equipment, transforming drones, unmanned vessels, and underwater robots from isolated operational units into a fully interconnected organic whole. This model not only dynamically adjusts inspection strategies based on real-time environmental changes to achieve a balance between safety risks and resource efficiency, but also continuously improves prediction accuracy through iterative historical data iteration. It fundamentally solves the inherent shortcomings of traditional technologies—the lack of cross-domain collaborative mechanisms and difficulty in adapting to complex dynamic environments—promoting inspection management in scenarios such as offshore wind farms from a fragmented approach to a new level of intelligent integration.

[0096] Those skilled in the art will recognize that the units of the various examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of the invention.

[0097] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical functional division. In actual implementation, there may be other division methods, such as multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored.

[0098] Furthermore, the functional units in the various embodiments of the present invention 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.

[0099] 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 the present invention, 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.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0100] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A method for integrated sea, land, air, and underwater inspection and dispatching, characterized in that, Includes the following steps: Collect and organize historical meteorological data, equipment failure records, UAV flight logs, power consumption data, and task assignment records; Assess the potential impact of severe weather and equipment failure risks that different regions may face when carrying out future missions, calculate the comprehensive risk value for all mission areas, and classify the risk levels. A scheduling optimization model is constructed with the goal of minimizing the overall weighted cost. The risk prediction results are input into the scheduling optimization model to solve and generate the optimal UAV scheduling scheme. The expression of the objective function is as follows: In the formula, It is the total weighted average cost; Is the type as A collection of drones These are the weighting coefficients for economic cost, time cost, energy cost, and risk aversion cost, respectively, and they are added together to equal 1. For economic costs; For time cost; For energy consumption costs, To mitigate the costs of risk.

2. The integrated sea, land, air, and underwater inspection and dispatch method according to claim 1, characterized in that, The process of risk level classification is as follows: Obtain severe weather risk assessment indicators, including at least the frequency of wind speed exceeding limits, the frequency of rainfall exceeding limits, the frequency of visibility falling below the threshold, and the frequency of lightning activity; Obtain equipment failure risk assessment indicators, including at least equipment failure rate, mean time between failures, and mean time to repair. The entropy value and weight of each risk indicator are calculated, and then the overall risk value is calculated, as follows: In the formula, Indicates the first i The overall risk value for each region; Indicates the first j The weight of each indicator; Indicates the first i The region in the first j Standardized values ​​for each risk indicator; The calculated comprehensive risk value They are divided into different risk levels.

3. The integrated sea, land, air, and underwater inspection and dispatch method according to claim 1, characterized in that, The scheduling optimization model also includes constraints, including at least the number of drones: the number of drones performing tasks simultaneously cannot exceed the total number of available drones.

4. The integrated sea, land, air, and underwater inspection and dispatch method according to claim 1, characterized in that, The scheduling optimization model also includes constraints, including at least the task uniqueness constraint: each task area must be executed by one drone only once.

5. The integrated sea, land, air, and underwater inspection and dispatch method according to claim 1, characterized in that, The scheduling optimization model also includes constraints, including at least a drone type matching constraint: task allocation must take into account the applicability of drone types.

6. The integrated sea, land, air, and underwater inspection and dispatch method according to claim 1, characterized in that, The scheduling optimization model also includes constraints, including at least the drone's endurance constraint: the total energy consumption of the drone during mission execution cannot exceed its battery capacity.

7. The integrated sea, land, air, and underwater inspection and dispatch method according to claim 1, characterized in that, It also includes the following steps: Based on historical data and future task predictions, a charging station site selection optimization model is run to determine the optimal location and number of charging stations. The specific details of the charging station site selection optimization model are as follows: In the formula, This represents the comprehensive benefit index, which is the ultimate goal of model optimization. This indicates the task coverage rate, which is the proportion of the number of task areas served by the charging station to the total number of task areas. This represents the total cost, including the construction cost and long-term operating cost of the charging station; The risk cost is mainly considered in terms of the risk of mission failure due to unreasonable charging station layout and the risk of the drone running out of power. These are the weighting coefficients for task coverage, total cost, and risk cost, respectively.

8. The integrated sea, land, air, and underwater inspection and dispatch method according to claim 7, characterized in that, The charging station site selection optimization model also includes constraints, including at least a task coverage constraint: each task area must be covered by at least one charging station.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein, when the program is executed, it controls the device containing the computer-readable storage medium to perform the integrated sea, land, air, and underwater inspection and scheduling method according to any one of claims 1 to 8.

10. A processor, characterized in that, The processor is used to run a program, wherein the program executes the integrated sea, land, air and submarine inspection and scheduling method according to any one of claims 1 to 8.

Images