Unmanned aerial vehicle task online matching system based on multi-dimensional evaluation and space-time constraints

Abstract

This invention relates to the field of UAV mission scheduling technology and discloses an online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints. The system includes a mission urgency assessment module, a UAV suitability calculation module, and an online matching optimization decision module. The mission urgency assessment module acquires attribute data of the set of tasks to be matched and calculates the multidimensional spatiotemporal coupling urgency of each task. The UAV suitability calculation module acquires the state data of the online UAV set and calculates the suitability value between each UAV and each task based on the multidimensional spatiotemporal coupling urgency. The online matching optimization decision module constructs a multi-objective optimization model based on the suitability values ​​and solves it to obtain the mission matching scheme. This invention introduces a mechanism for evaluating the spatiotemporal coupling effect between tasks, improving the global optimization level of matching in multi-task scenarios.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) mission scheduling technology, and in particular to an online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints. Background Technology

[0002] Existing UAV mission matching systems mostly adopt a single-mission independent evaluation mode, analyzing the attributes of each mission separately before making matching decisions. This evaluation mode can meet basic needs in scenarios with a small number of missions and low correlation between missions, but it has significant limitations in large-scale multi-mission concurrent scenarios. Because it does not consider the coupling and correlation effects between missions in the temporal and spatial dimensions, it cannot accurately reflect the overall urgency and resource demand distribution of the mission cluster, resulting in matching decisions that can only achieve local optimization rather than global optimization, thus restricting mission execution efficiency and resource utilization. With the continuous expansion of UAV application scenarios and the continuous increase in mission complexity, there is an urgent need for a mission matching technology solution that can comprehensively consider the spatiotemporal coupling relationships between missions. Summary of the Invention

[0003] The purpose of this invention is to address the shortcomings of existing technologies by proposing an online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints. This system includes a mission urgency assessment module, a UAV suitability calculation module, and an online matching optimization decision module. The mission urgency assessment module is connected to the UAV suitability calculation module and is used to acquire attribute data of the set of tasks to be matched and calculate the multidimensional spatiotemporal coupling urgency of each task. The UAV suitability calculation module is connected to the online matching optimization decision module and is used to acquire state data of the online UAV set and calculate the suitability value between each UAV and each task to be matched based on the multidimensional spatiotemporal coupling urgency. The online matching optimization decision module is used to construct a multi-objective optimization model based on the suitability values ​​and solve the multi-objective optimization model to obtain a mission matching scheme.

[0004] Preferably, the task urgency assessment module includes a task data acquisition unit, a urgency calculation unit, and a coupling calculation unit. The task data acquisition unit is connected to the urgency calculation unit and is used to acquire deadline data, coverage area data, resource requirement data, task type data, and geographic location coordinate data for each task to be matched. The urgency calculation unit is connected to the coupling calculation unit and is used to generate a time urgency component based on the deadline data, a spatial urgency component based on the coverage area data, and a resource urgency component based on the resource requirement data. The coupling calculation unit is used to calculate the coupling effect strength between each task to be matched based on the task type data and the geographic location coordinate data, and to generate the multidimensional spatiotemporal coupling urgency based on the time urgency component, the spatial urgency component, and the resource urgency component.

[0005] More preferably, the UAV adaptability calculation module includes a UAV status acquisition unit, a single-factor evaluation unit, and a synergistic effect calculation unit; the UAV status acquisition unit is connected to the single-factor evaluation unit and is used to collect the current location data, remaining battery power data, payload configuration data, total battery capacity data, and health status data of each online UAV; the single-factor evaluation unit is connected to the synergistic effect calculation unit and is used to generate a distance reachability index based on the current location data, a battery adequacy index based on the remaining battery power data and the total battery capacity data, and a payload matching index based on the payload configuration data; the synergistic effect calculation unit is used to generate a synergistic gain value based on the health status data and assigned task information, and to generate the adaptability value based on the distance reachability index, the battery adequacy index, and the payload matching index.

[0006] More preferably, the online matching optimization decision module includes an objective function construction unit, a constraint processing unit, and an optimization solution unit. The objective function construction unit is connected to the constraint processing unit and is used to construct a matching benefit term based on the fitness value, a non-matching penalty term based on the multi-dimensional spatiotemporal coupling urgency, and a load balancing penalty term based on the load distribution data of each UAV. The objective function is generated based on the matching benefit term, the non-matching penalty term, and the load balancing penalty term. The constraint processing unit is connected to the optimization solution unit and is used to generate UAV task allocation constraints, task execution capability constraints, and time window constraints. The optimization solution unit is used to solve the multi-objective optimization model based on the objective function, the UAV task allocation constraints, the task execution capability constraints, and the time window constraints, and output the task matching scheme.

[0007] Further preferably, the coupling calculation unit includes an iteration control subunit and a convergence judgment subunit; the iteration control subunit is connected to the convergence judgment subunit and is used to sequentially read the urgency data and surrounding task information of each task to be matched according to a preset iteration order, calculate the coupling correction urgency of each task to be matched, and write the coupling correction urgency into a storage area; the convergence judgment subunit is used to read the urgency data of each task to be matched in two adjacent iterations, calculate the sum of the urgency differences between the two iterations, and generate an iteration termination signal when the sum of the urgency differences is less than the convergence threshold and output it to the iteration control subunit.

[0008] Further preferably, the single-factor evaluation unit further includes a weight adaptive adjustment subunit; the weight adaptive adjustment subunit is connected to the synergistic effect calculation unit, and is used to acquire the current system's task quantity distribution data, task type distribution data, drone quantity distribution data, and drone status distribution data, generate distance dimension weight coefficients, battery dimension weight coefficients, and payload dimension weight coefficients based on the task quantity distribution data, task type distribution data, drone quantity distribution data, and drone status distribution data, and output the distance dimension weight coefficients, battery dimension weight coefficients, and payload dimension weight coefficients to the synergistic effect calculation unit.

[0009] Further preferably, the optimization solution unit includes a scale judgment subunit, an exact solution subunit, and a heuristic solution subunit; the scale judgment subunit is connected to the exact solution subunit and the heuristic solution subunit, and is used to count the number of tasks to be matched and the number of online drones, compare the number of tasks to be matched and the number of online drones with a scale threshold respectively, and select the exact solution subunit or the heuristic solution subunit to perform the solution operation based on the comparison result; the exact solution subunit is used to solve the multi-objective optimization model using an integer programming algorithm when the number of tasks to be matched and the number of online drones are both less than the scale threshold; the heuristic solution subunit is used to solve the multi-objective optimization model using a heuristic search algorithm when the number of tasks to be matched or the number of online drones is greater than or equal to the scale threshold.

[0010] Further preferably, the system also includes a scrolling window rematching module; the scrolling window rematching module is connected to the task urgency assessment module and the UAV adaptability calculation module, and includes a status monitoring subunit, a trigger judgment subunit, and an incremental calculation subunit; the status monitoring subunit is connected to the trigger judgment subunit and is used to monitor task addition events, task cancellation events, and task attribute change events on the task side, as well as UAV failure events, UAV battery mutation events, and UAV payload change events on the UAV side, and generate a status change notification signal; the trigger judgment subunit is connected to the incremental calculation subunit and is used to generate a rematching trigger signal according to the status change notification signal and the rematching triggering rules; the incremental calculation subunit is used to determine the task set and UAV set affected by the status change according to the rematching trigger signal, perform the update operation of the multidimensional spatiotemporal coupling urgency and the adaptability value on the task set and UAV set affected by the status change, and output the update result to the online matching optimization decision module.

[0011] Further preferably, the system also includes an execution feedback and parameter tuning module; the execution feedback and parameter tuning module is connected to the task urgency assessment module and the UAV adaptability calculation module, and includes an execution data acquisition subunit, a deviation calculation subunit, and a parameter adaptive adjustment subunit; the execution data acquisition subunit is connected to the deviation calculation subunit and is used to collect actual energy consumption data, actual completion time data, and actual completion quality data during the actual execution of each task; the deviation calculation subunit is connected to the parameter adaptive adjustment subunit and is used to calculate the energy consumption deviation value between the actual energy consumption data and the estimated energy consumption data, calculate the time deviation value between the actual completion time data and the estimated completion time data, and calculate the quality deviation value between the actual completion quality data and the estimated completion quality data; the parameter adaptive adjustment subunit is used to generate urgency model parameter adjustment signals and adaptability model parameter adjustment signals based on the energy consumption deviation value, the time deviation value, and the quality deviation value, output the urgency model parameter adjustment signals to the task urgency assessment module, and output the adaptability model parameter adjustment signals to the UAV adaptability calculation module.

[0012] Further preferably, the system also includes a spatiotemporal constraint verification module; the spatiotemporal constraint verification module is connected to the task urgency assessment module and the online matching optimization decision module, and is used to acquire time window data and spatial region data of each task to be matched, calculate the degree of spatiotemporal conflict between tasks based on the time window data and the spatial region data, generate a spatiotemporal constraint matrix, and output the spatiotemporal constraint matrix to the online matching optimization decision module; the online matching optimization decision module is also used to solve the multi-objective optimization model in combination with the spatiotemporal constraint matrix.

[0013] Technical Effects: This invention calculates the multi-dimensional spatiotemporal coupling urgency of each task to be matched through a task urgency assessment module, introducing an assessment mechanism for the spatiotemporal coupling effect between tasks, thus solving the problem of insufficient matching accuracy caused by independent task assessment in existing technologies. The UAV adaptability calculation module combines urgency and multi-dimensional state data to calculate adaptability values, and the online matching optimization decision module constructs a multi-objective optimization model to solve the matching scheme, achieving an improvement in task matching from local optimum to global optimum, thereby improving matching quality and resource utilization efficiency in multi-task scenarios. Attached Figure Description

[0014] Figure 1 This is a connection block diagram of the online matching system for UAV missions based on multidimensional evaluation and spatiotemporal constraints proposed in this application;

[0015] Figure 2 This is a schematic diagram showing the spatial distribution and matching connections between the mission and the UAV in this application;

[0016] Figure 3 A visualization of the spatiotemporal matching effect for multi-UAV mission allocation in this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] Existing technologies independently evaluate tasks without considering the interactions between tasks, resulting in significant matching biases.

[0019] Based on this, please refer to Figures 1-3 This embodiment provides an online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints. The system includes a mission urgency assessment module, a UAV suitability calculation module, and an online matching optimization decision module. The mission urgency assessment module is connected to the UAV suitability calculation module, acquiring attribute data of the set of tasks to be matched and calculating the multidimensional spatiotemporal coupling urgency of each task. The UAV suitability calculation module is connected to the online matching optimization decision module, acquiring state data of the online UAV set and calculating the suitability value between each UAV and each task to be matched based on the multidimensional spatiotemporal coupling urgency. The online matching optimization decision module constructs a multi-objective optimization model based on the suitability values ​​and solves the multi-objective optimization model to obtain a mission matching scheme.

[0020] The task urgency assessment module, serving as the system's input assessment unit, comprehensively evaluates the urgency of all tasks to be matched. This module assesses tasks from three dimensions: time, space, and resources, while also incorporating inter-task interactions to ensure the assessment results reflect the relative urgency of tasks within the entire task cluster. This multi-dimensional spatiotemporal coupling urgency serves as the foundational input for subsequent suitability calculations and matching decisions, directly impacting the quality of the final matching scheme. The UAV suitability calculation module, acting as an intermediate calculation unit, assesses the suitability of each UAV for each task. The assessment process combines the UAV's own state with the task requirements, along with the task's urgency level, ensuring that high-urgency tasks receive higher matching priority. The online matching optimization decision module, as the system's core decision-making unit, transforms the matching problem into a multi-objective optimization problem and solves it, comprehensively considering matching benefits, non-matching penalties, and load balancing to achieve the globally optimal matching result.

[0021] The three modules form a progressive evaluation chain. The output of the task urgency evaluation module serves as the input of the UAV suitability calculation module, and the output of the UAV suitability calculation module serves as the input of the online matching optimization decision module. The data flow is clear, and the logical relationships are well-defined. This hierarchical and progressive evaluation architecture allows each module to focus on calculations at a specific level, ensuring both the comprehensiveness of the evaluation and facilitating independent optimization and maintenance of each module.

[0022] The task urgency assessment module includes a task data acquisition unit, a urgency calculation unit, and a coupling calculation unit. The task data acquisition unit is connected to the urgency calculation unit and collects deadline data, coverage area data, resource requirement data, task type data, and geographic location coordinate data for each task to be matched. The urgency calculation unit is connected to the coupling calculation unit and generates a time urgency component based on the deadline data, a spatial urgency component based on the coverage area data, and a resource urgency component based on the resource requirement data. The coupling calculation unit calculates the coupling effect strength between each task to be matched based on the task type data and the geographic location coordinate data, and generates the multidimensional spatiotemporal coupling urgency based on the time urgency component, the spatial urgency component, and the resource urgency component.

[0023] The task data acquisition unit obtains complete attribute information of all tasks to be matched from the task scheduling platform or task input interface, providing a data foundation for subsequent evaluation. The acquired data covers five aspects: time attributes, spatial attributes, resource attributes, type attributes, and location attributes, ensuring that subsequent evaluation can be carried out from multiple dimensions. The urgency calculation unit converts the raw data of each dimension into urgency components with a unified dimension, each component corresponding to an evaluation dimension. The time urgency component reflects the urgency of the task in the time dimension; the closer the deadline, the higher the time urgency. The spatial urgency component reflects the coverage or operational complexity of the task in the spatial dimension; the larger the coverage area, the higher the spatial urgency. The resource urgency component reflects the degree of resource demand of the task; the greater the resource demand, the higher the resource urgency. The three components characterize the urgency of the task from different perspectives, avoiding the one-sidedness of single-dimensional evaluation. The coupling calculation unit calculates the coupling effect between tasks and incorporates the coupling effect into the final urgency calculation result. The calculation of the coupling effect is based on the task type correlation and spatial distance attenuation relationship; tasks with similar types and close spatial distances have a stronger coupling effect.

[0024] The specific calculation of the multidimensional spatiotemporal coupling urgency is achieved through the task multidimensional spatiotemporal coupling urgency formula, which is expressed as:

[0025]

[0026] The left side of the formula The urgency of the multidimensional spatiotemporal coupling of the j-th task is represented by a dimensionless value. Since urgency is a relative evaluation value used for comparison and ranking between different tasks, using a dimensionless normalized numerical representation is most reasonable and facilitates subsequent calculations and comparisons. The right side of the formula is divided into two parts: the weighted power sum part within square brackets and the coupling amplification part within brackets. Multiplying the two parts yields the final urgency value.

[0027] The part within the square brackets is the weighted sum of the three urgency components. , , These represent the basic weights for the time, space, and resource dimensions, respectively. All three are dimensionless. These three weight coefficients are used to adjust the relative importance of different dimensions in the overall evaluation; a larger weight value indicates a greater impact of the corresponding dimension on the final urgency level. The values ​​of the three weight coefficients are typically set between 0 and 1, and their sum can be set according to actual needs. It can be set to a normalized sum of 1, or other numerical combinations. By adjusting the weight ratio, the importance differences of each dimension in different scenarios can be adapted. , , These represent the time urgency, space urgency, and resource urgency components of task j, respectively. All components are dimensionless, and each component typically ranges from 0 to 1, with higher values ​​indicating greater urgency in the corresponding dimension. These three components characterize the urgency attributes of the task from different perspectives, providing foundational data for subsequent multi-dimensional fusion.

[0028] , , These represent the nonlinear adjustment indices for the three dimensions, all dimensionless. These indices are used to adjust the nonlinear mapping relationship of each dimension's components. An index greater than 1 will amplify the high-value range of the component, while an index less than 1 will compress the high-value range. Adjusting the indices can change the discriminative power between tasks of different urgency levels. The index values ​​are typically set between 0.5 and 2, with the specific value determined based on the changing characteristics of urgency in each dimension within the actual scenario. For example, when the urgency of the time dimension needs to increase rapidly near the deadline, the index can be adjusted... Setting it to a value greater than 1 amplifies the high-value range of the time urgency component, enhancing the differentiation of urgency for tasks nearing their deadlines. (The square brackets are used to denote the overall urgency level.) The exponentiation operation is used to normalize the weighted sum of powers, ensuring that the output value remains within a reasonable range and dimensionless regardless of changes in the weights and exponents of the three components. This normalization method borrows from the calculation idea of ​​Minkowski distance, using the reciprocal of the exponent sum as a normalization factor to keep the order of magnitude of the output value relatively stable under different parameter settings.

[0029] The second part on the right side of the formula is the coupling effect amplification factor, in the form of: . This represents the coupling effect strength coefficient, which is dimensionless and used to adjust the degree of influence of the coupling effect on the final urgency. The larger the coefficient, the more significant the effect of the coupling effect of surrounding tasks on the urgency of the current task. The value of the coupling effect strength coefficient is usually set between 0 and 2. A value of 0 indicates that the coupling effect is not considered at all, while a larger value indicates that the influence of the coupling effect is stronger. The specific value needs to be determined based on the coupling characteristics of the task cluster. This represents the degree of coupling of task j to its surrounding tasks. It is dimensionless and reflects the sum of the coupling effects of all surrounding tasks on the current task. When the degree of coupling is 0, the amplification factor equals 1, meaning there is no coupling effect. When the degree of coupling is greater than 0, the amplification factor is greater than 1, meaning the presence of surrounding tasks increases the urgency of the current task. This design aligns with the real-world scenario where the overall scheduling pressure increases when multiple similar tasks cluster together. The scheduling resource requirements of task clusters are higher than the simple sum of scattered tasks, and this clustering effect can be quantified through the coupling amplification factor.

[0030] The formula's theoretical design is based on multi-attribute decision theory and field theory. The three urgency components are fused using a weighted power sum, a common weighted aggregation method in multi-attribute decision-making. The power exponent allows for flexible adjustment of the nonlinear contributions of each dimension. The coupling effect part borrows from the superposition concept in physical fields, treating each task as a field source that influences surrounding tasks. The intensity of this influence decays with distance and is type-dependent. This design ensures that urgency assessment considers not only the task's own attributes but also the cluster environment in which the task exists, better meeting the needs of real-world scheduling scenarios.

[0031] The formula for calculating the time urgency component is as follows: ,in This represents the deadline for task j, measured in units of time. This indicates the current system time, measured in units of time. The time constant, representing the urgency of time, is measured in units of time. (Molecular term) This represents the remaining time for the task, measured in units of time. Dividing it by the characteristic time constant makes it dimensionless and serves as the input to the exponential function. The exponential function uses a negative exponential form. When the remaining time is long, the exponent term approaches 0, and the time urgency component approaches 1 minus 0 equals 1; when the remaining time approaches 0, the exponent term approaches 1, and the time urgency component approaches 0. Smaller values ​​indicate greater time urgency. This relationship can be accurately reflected in the final result through adjustments to weights and the exponent in the overall formula. The exponential function form is used because the change in time urgency typically exhibits non-linear characteristics; the closer to the deadline, the faster the urgency increases, and the exponential function effectively simulates this characteristic. Characteristic time constant. It controls the rate at which the urgency level changes with the remaining time. The smaller the value, the steeper the change in urgency. The larger the value, the more gradual the change in urgency. The specific value needs to be determined based on the time-sensitive characteristics of the task.

[0032] The formula for calculating the spatial urgency component is as follows: ,in This represents the area covered by task j, with the dimension being the square of the length. The characteristic area constant represents the spatial urgency, with dimensions of length squared. The area ratio is dimensionless and serves as the input to the exponential function. When the task coverage area is small, the exponential term approaches 0, and the spatial urgency component approaches 0; as the coverage area increases, the spatial urgency component monotonically increases and gradually approaches 1. This design reflects the characteristic that larger coverage areas indicate higher task spatial complexity and urgency, with the growth rate gradually slowing down, conforming to the law of diminishing marginal returns. Characteristic area constant It controls the rate at which the spatial urgency changes with area. The smaller the value, the faster the urgency level increases. The larger the value, the more gradual the increase in urgency. The specific value needs to be determined based on the spatial scale characteristics of the task.

[0033] The formula for calculating the resource urgency component is as follows: ,in This represents the total resource requirement for task j, measured in units of resource quantity. This represents the reference threshold for resource demand, measured in terms of resource quantity. The resource demand ratio is dimensionless and serves as the input to the hyperbolic tangent function. The hyperbolic tangent function ranges from -1 to 1; when the input is positive, the output is between 0 and 1, gradually approaching 1 as the input increases. The hyperbolic tangent function is used because resource urgency increases rapidly in the low-demand range and tends to level off in the high-demand range, and the function has favorable mathematical properties, facilitating subsequent differentiation and optimization calculations. Resource Demand Reference Threshold It controls the inflection point of resource urgency growth. When resource demand approaches the threshold, urgency increases rapidly. After exceeding the threshold, the growth gradually slows down. The specific value needs to be determined based on the system's resource supply capacity.

[0034] The formula for calculating the peripheral coupling degree is: ,in This represents the type correlation coefficient between task j and task k. It is dimensionless and its value usually ranges from 0 to 1. The more similar the types are, the larger the coefficient. This represents the spatial distance between task j and task k, with the dimension being length. The spatial coupling attenuation characteristic scale is represented by length. The ratio of distance to the characteristic scale is dimensionless, and the negative exponential function indicates that the coupling effect decays with distance; the coupling effect weakens with increasing distance. The peripheral coupling degree is calculated by summing the coupling contributions of all other tasks. Each task's contribution is obtained by multiplying its urgency, type correlation coefficient, and distance attenuation factor. This product form reflects the coupling effect of multiple factors working together. Spatial Coupling Attenuation Characteristic Scale It controls the rate at which the coupling effect decays with distance. The smaller the value, the faster the decay. The larger the value, the slower the decay; the specific value needs to be determined based on the spatial influence range of the task.

[0035] The UAV adaptability calculation module includes a UAV status acquisition unit, a single-factor evaluation unit, and a synergistic effect calculation unit. The UAV status acquisition unit is connected to the single-factor evaluation unit and collects current location data, remaining battery power data, payload configuration data, total battery capacity data, and health status data for each online UAV. The single-factor evaluation unit is connected to the synergistic effect calculation unit and generates a distance reachability index based on the current location data, a battery adequacy index based on the remaining battery power data and the total battery capacity data, and a payload matching index based on the payload configuration data. The synergistic effect calculation unit generates a synergistic gain value based on the health status data and assigned task information, and generates the adaptability value based on the distance reachability index, the battery adequacy index, and the payload matching index.

[0036] The UAV status acquisition unit obtains real-time status data from the onboard communication modules or ground control stations of each UAV, providing data support for adaptability assessment. The acquired data covers five aspects: location, battery level, payload, total capacity, and health status, ensuring that subsequent assessments can measure the UAV's suitability for the mission from multiple dimensions. The single-factor assessment unit calculates adaptability indicators for each dimension, with each indicator corresponding to an assessment dimension. The distance accessibility indicator reflects the impact of the distance from the UAV's current location to the mission location on mission execution; the closer the distance, the higher the accessibility. The battery adequacy indicator reflects the remaining battery level of the UAV after completing the mission; the more battery remaining, the higher the battery adequacy and the stronger the ability to perform subsequent tasks. The payload matching indicator reflects the degree of matching between the UAV's payload configuration and mission requirements; the better the configuration, the higher the matching degree. These three indicators characterize the UAV's adaptability to the mission from different perspectives. The synergy effect calculation unit calculates the synergy gain of multi-task combined execution and incorporates the synergy gain into the final adaptability calculation. When a drone has already been assigned other tasks, there may be synergies between the new task and the assigned task. For example, if the task locations are close, it can reduce the need for relocation flights, or if the task types are similar, they can share some payload configurations. These will improve the efficiency of combined execution.

[0037] The specific calculation of the adaptability value is achieved through the UAV mission adaptability formula, which is expressed as:

[0038]

[0039] The left side of the formula This represents the fitness score of the i-th UAV for the j-th task, and its dimensionless value. Fitness score is a relative evaluation value used to compare and rank different UAVs with different task combinations; therefore, it is represented by a dimensionless normalized numerical value. The right side of the formula consists of three parts: the UAV state adjustment coefficient, the weighted sum part, and the cooperative gain part. Multiplying these three parts yields the final fitness score value.

[0040] This represents the state adjustment coefficient for the i-th drone. It is dimensionless and used to adjust the overall fit based on the drone's overall health status or performance level. A higher coefficient indicates a better drone's condition and thus a higher overall fit. The state adjustment coefficient is typically set between 0.5 and 1.5. A drone in normal condition has a coefficient of 1, a drone in better condition has a coefficient greater than 1, and a drone in worse condition has a coefficient less than 1. Adjusting this coefficient reflects the performance differences between different drones.

[0041] The part within square brackets is the weighted sum of the four evaluation factors. , , , These represent the weighting coefficients for power consumption, urgency matching, load matching, and distance factors, respectively. All four coefficients are dimensionless and are used to adjust the relative importance of different factors in the overall fit. The values ​​of these four coefficients are typically set between 0 and 1, but the specific values ​​need to be determined based on the importance of each factor in the matching decision. Adjusting the weighting ratios can adapt to matching preferences in different scenarios.

[0042] This represents the normalized remaining battery power after completing a task. It is dimensionless; a larger value indicates more battery power remaining, meaning a more abundant battery. The formula for calculating the normalized remaining battery power is: ,in This represents the current remaining battery power of drone i, expressed in units of energy. This represents the estimated energy consumption of drone i performing task j, with the dimension being energy; The total battery capacity of drone i is expressed in terms of energy. The numerator is the remaining battery capacity minus the estimated energy consumption, yielding the remaining battery capacity after the mission, also expressed in terms of energy. Dividing this by the total battery capacity results in a dimensionless, normalized value. This calculation method accurately reflects the drone's battery level after the mission; the more battery capacity reserved for subsequent missions, the higher the adaptability. The estimated energy consumption can be determined based on mission distance, mission duration, and the drone's energy consumption characteristics, ensuring that the deviation between the estimated result and the actual energy consumption is within an acceptable range.

[0043] This indicates a matching item based on urgency, and its dimension is dimensionless. It is the multidimensional spatiotemporal coupling urgency of task j. This is the normalized benchmark for the system's maximum urgency; both are dimensionless, and the ratio is also dimensionless. The expression for this term is 1 minus the urgency ratio, meaning the higher the urgency of the task, the smaller this term's value. In the weighted sum of adaptability, this term represents the matching relationship between task urgency and drone capabilities. High-urgency tasks place higher demands on drones, and the adaptability of ordinary drones will decrease accordingly. From another perspective, this term reflects the drone's ease in handling high-urgency tasks; the lower the urgency of the task, the more easily the drone can execute it, and the higher its adaptability. This is determined by weighting coefficients. The positive and negative adjustments can flexibly control the direction of this relationship, when When the value is positive, tasks with low urgency have a higher suitability. When a negative value is used, tasks with higher urgency have a higher compatibility. The specific value needs to be determined based on the matching strategy.

[0044] This represents the payload matching degree between UAV i and task j. It is dimensionless and typically ranges from 0 to 1. A higher value indicates that the UAV's payload configuration better meets the task's payload requirements, resulting in a higher matching degree. The payload matching degree can be determined based on the degree of matching between the UAV's payload type, payload capacity, and the task's payload requirements. A matching degree of 1 indicates that the UAV's payload fully meets the task requirements, while a matching degree of 0 indicates that the UAV's payload does not meet the task requirements at all.

[0045] This represents the distance attenuation term, which is dimensionless. This represents the spatial distance from drone i to mission j, with the dimension being length. This represents the distance decay characteristic scale parameter, with dimensions of length. The ratio of distance to the characteristic scale is dimensionless. The negative exponential function represents the characteristic of fitness decaying with distance; the fitness decreases with increasing distance. The exponential decay form is used because the effect of distance on fitness is usually non-linear, with a larger impact at close range and a gradually weakening impact at greater distances. The exponential function can well simulate this characteristic. Distance decay characteristic scale parameter It controls the rate at which fitness decreases with distance. The smaller the value, the faster the decay. The larger the value, the slower the decay; the specific value needs to be determined based on the distance-sensitive characteristics of the task.

[0046] The third part on the right side of the formula is the synergistic gain factor, in the form of: . This represents the synergy strength coefficient, which is dimensionless and used to adjust the degree of influence of the synergy effect on the final fit. The value of the synergy strength coefficient is usually set between 0 and 1. A value of 0 indicates that the synergy effect is not considered at all, while a larger value indicates that the influence of the synergy effect is stronger. The specific value needs to be determined based on the actual gain level of multi-task collaboration. This represents the degree of synergy between task j and the assigned tasks of drone i. It is dimensionless and reflects the magnitude of the synergistic gain from the combined execution of the new task and the assigned tasks. When the synergy is 0, the gain factor equals 1, meaning there is no synergistic effect; when the synergy is greater than 0, the gain factor is greater than 1, meaning that the combined execution of multiple tasks will improve the overall adaptability.

[0047] The formula for calculating the degree of synergy is: ,in The variable represents the matching decision between drone i and task l. It is dimensionless and takes the value 0 or 1, where 1 indicates a match and 0 indicates no match. This represents the synergy weight between task j and task l, with dimensionless dimensions, reflecting the degree of synergy between the two tasks. The synergy degree is calculated by summing the synergy contributions of all assigned tasks. Each task's contribution is obtained by multiplying its fitness, matching status, and synergy weight, reflecting the cumulative synergy effect of multi-task combinations. The synergy weight is typically set between 0 and 1; the stronger the synergy between two tasks, the greater the weight. The synergy weight can be determined based on a comprehensive calculation of multiple factors such as task type similarity, spatial proximity, and temporal overlap.

[0048] The formula's theoretical design is based on multi-attribute utility theory and synergy theory. The four evaluation factors are integrated using a weighted sum, a standard form for multi-attribute utility evaluation. Weighting coefficients allow for flexible adjustment of the relative importance of each factor. The synergy gain component embodies the idea of ​​combinatorial optimization; multi-task combinations may produce a synergistic effect greater than the sum of its parts. The synergy gain factor quantifies this effect and incorporates it into the fit calculation, enabling matching decisions to consider the overall benefits of multi-task combinations.

[0049] The online matching optimization decision module includes an objective function construction unit, a constraint processing unit, and an optimization solution unit. The objective function construction unit is connected to the constraint processing unit. It constructs a matching benefit term based on the fitness value, a non-matching penalty term based on the multi-dimensional spatiotemporal coupling urgency, and a load balancing penalty term based on the load distribution data of each UAV. The objective function is generated based on the matching benefit term, the non-matching penalty term, and the load balancing penalty term. The constraint processing unit is connected to the optimization solution unit and generates UAV task allocation constraints, task execution capability constraints, and time window constraints. The optimization solution unit solves the multi-objective optimization model based on the objective function, the UAV task allocation constraints, the task execution capability constraints, and the time window constraints, and outputs the task matching scheme.

[0050] The objective function construction unit combines the various objective terms of the multi-objective optimization problem into a unified objective function. The matching benefit term reflects the overall fitness benefit of the matching scheme; the higher the fitness of all matching pairs, the greater the total benefit. The unmatched penalty term reflects the loss caused by unmatched tasks; the higher the urgency of a task, the greater the penalty when it fails to match, prompting high-urgency tasks to be matched first. The load balancing penalty term reflects the balance of the load distribution among the drones; the more unbalanced the load, the greater the penalty, prompting tasks to be evenly distributed among the drones. These three objective terms are combined with weights to form a unified objective function; adjusting the weights can balance the relative importance of different objectives. The constraint processing unit transforms various constraints of the matching problem into mathematical constraints. Drone task allocation constraints ensure that the number of tasks assigned to each drone is within a reasonable range. Task execution capability constraints ensure that the drones have the ability to execute the corresponding tasks. Time window constraints ensure that tasks can be completed within a specified time window. The optimization solution unit, under the premise of satisfying all constraints, solves for the optimal solution of the objective function to obtain the final task matching scheme.

[0051] The specific calculation of the total benefit of system matching is achieved through the formula for the total benefit of system matching, which is expressed as follows:

[0052]

[0053] The left side of the formula This represents the total benefit value of the system matching, and its dimensionless value. The total benefit value is a comprehensive indicator for measuring the quality of a matching scheme; the larger the value, the better the overall benefit of the matching scheme. The right side of the formula is divided into three parts: matching benefit, non-matching penalty, and load balancing penalty, which are combined through addition and subtraction to obtain the total benefit.

[0054] The first part of the matching benefit items is: The dimensionless quantity. It is the degree of adaptability of drone i to task j, which is dimensionless; This is the matching decision variable, dimensionless, and takes the value of 0 or 1. Multiplying these two variables and summing them over all drones and all tasks yields the sum of the fitness scores for all matching pairs, i.e., the total matching reward. This represents the total number of drones, and is a dimensionless count. This represents the total number of tasks to be matched, measured as a dimensionless count. Both are count variables and do not participate in dimensional calculations. The matching benefit term reflects the overall quality of the matching scheme; the higher the fit of all matching pairs, the greater the total matching benefit, and the better the matching scheme. Matching decision variables. The value of is determined by the optimization solution process. When UAV i matches task j, the variable takes the value 1, and when they do not match, the variable takes the value 0. The matching relationship is transformed into a mathematical form by taking the values ​​of 0-1 of the variable.

[0055] The second part of the non-matching penalty items is as follows: The dimensionless quantity. This represents the penalty coefficient for unmatched tasks. It is dimensionless and used to adjust the intensity of the penalty. The value of the penalty coefficient for unmatched tasks is usually set between 0 and 2. The specific value needs to be determined based on the tolerance level for unmatched tasks. The larger the coefficient, the heavier the penalty for unmatched tasks, and the more inclined it is to allow more tasks to be matched. It represents the multidimensional spatiotemporal coupling urgency of task j, which is dimensionless. (The part in parentheses...) This represents the unmatched state of task j, with dimensionless dimensions. When task j is not matched by any drone, the summation result is 0 (with 1 inside the parentheses); when task j is matched by a drone, the summation result is 1 (with 0 inside the parentheses). The urgency is multiplied by the unmatched state, and then summed over all tasks to obtain the total urgency penalty for unmatched tasks. Multiplying this by a penalty coefficient gives the unmatched penalty term. Subtracting this term from the total benefit means that unmatched tasks reduce the total benefit, with higher-urgency tasks experiencing a greater reduction when unmatched, thus encouraging high-urgency tasks to be matched first. This design reflects the correlation between task urgency and matching priority, ensuring that urgent tasks are processed first.

[0056] The third part is the load balancing penalty item. The dimensionless quantity. This represents the load balancing penalty coefficient, which is dimensionless and used to adjust the importance of load balancing. The load balancing penalty coefficient is usually set between 0 and 1, and the specific value needs to be determined according to the degree of load balancing requirements. The larger the coefficient, the heavier the penalty for load imbalance, and the more inclined it is to distribute tasks evenly among the drones. This represents the variance of the payload of each drone, with the dimension being the square of energy. The larger the variance, the more uneven the payload distribution. The average load level represents the overall system load, measured in units of energy. The square of the average load is also measured in units of energy squared. The ratio of variance to the square of the average load is dimensionless and called the square of the coefficient of variation, a commonly used indicator to measure relative dispersion. Relative dispersion is used instead of absolute variance because the importance of load balancing is related to the overall system load level, making a relative indicator more reasonable. Subtracting this term from the total benefit means that the more unbalanced the load, the lower the total benefit, thus promoting a more even distribution of tasks among the drones. Load balancing design can prevent some drones from being overloaded while others are underloaded, improving the overall system efficiency and stability.

[0057] The formula's theoretical design is based on multi-objective optimization theory and constrained optimization methods. The three objective terms correspond to maximizing matching rewards, minimizing unmatched penalties, and load balancing, respectively. A weighted summation transforms the multi-objective problem into a single-objective problem, facilitating its solution using mature optimization algorithms. Weight coefficients... and This is used to adjust the relative importance of different objectives and can be flexibly adjusted according to the needs of the actual scenario. It matches decision variables. The variables are 0-1, which makes the entire optimization problem an integer programming problem with a clear mathematical form and solution method.

[0058] The coupling calculation unit includes an iterative control subunit and a convergence judgment subunit. The iterative control subunit is connected to the convergence judgment subunit. It sequentially reads the urgency data and surrounding task information of each task to be matched according to a preset iteration order, calculates the coupling-corrected urgency of each task to be matched, and writes the coupling-corrected urgency into a storage area. The convergence judgment subunit reads the urgency data of each task to be matched in two adjacent iterations, calculates the sum of the urgency differences between the two iterations, and generates an iteration termination signal when the sum of the urgency differences is less than a convergence threshold, outputting it to the iterative control subunit.

[0059] The iterative control subunit controls the iterative process of the coupled computation, updating the urgency value of each task sequentially according to a preset order. In each iteration, the subunit reads the urgency data of all current tasks. For each task, it calculates the coupling effect based on the urgency and coupling relationship of its surrounding tasks, then corrects the urgency of the current task based on the coupling effect, obtaining the coupled-corrected urgency, and writes the corrected value back to the storage area. The iteration order can employ sequential iteration, random iteration, or other strategies. Different iteration strategies will affect the convergence speed and the final result, but will eventually converge to the same stable value. The convergence judgment subunit determines whether the iterative process has reached a convergence state. After each iteration, the subunit compares the urgency of all tasks in the current iteration with the result of the previous iteration, calculating the sum of the urgency differences of all tasks. When the sum of the differences is less than the convergence threshold, it indicates that the iteration result has stabilized and the change is small, reaching a convergence state. At this time, an iteration termination signal is generated, and the iteration process stops. If the sum of the differences is greater than or equal to the convergence threshold, the next round of iteration continues. The value of the convergence threshold needs to be determined based on the required computational accuracy. The smaller the threshold, the higher the computational accuracy, but the more iterations are required.

[0060] The single-factor evaluation unit further includes a weight adaptive adjustment subunit. This subunit is connected to the synergy effect calculation unit, acquires the current system's task quantity distribution data, task type distribution data, drone quantity distribution data, and drone status distribution data, generates distance dimension weight coefficients, battery level dimension weight coefficients, and payload dimension weight coefficients based on these data, and outputs these weight coefficients to the synergy effect calculation unit.

[0061] The adaptive weight adjustment subunit dynamically adjusts the weight coefficients of each dimension in the fit evaluation based on the current operating state of the system. The subunit first acquires various distribution data of the current system, including the distribution of the number of tasks, task types, number of drones, and drone status. This distribution data reflects the characteristics of the current scenario; for example, distance factors have a more prominent impact when the number of tasks is high, and battery factors have a more prominent impact when drone battery levels are generally low. Based on this distribution data, the subunit calculates the optimal weight coefficients for each dimension in the current scenario using preset weight adjustment rules or algorithms, and then outputs these weight coefficients to the synergy effect calculation unit. The principle of weight adjustment is that the dimension with the greatest impact on matching quality in the current scenario receives a higher weight. The frequency of weight adjustment can be set to be adjusted before each match or set to be adjusted periodically; the specific frequency needs to be determined based on the rate of change in the system state.

[0062] The optimization solution unit includes a scale judgment subunit, an exact solution subunit, and a heuristic solution subunit. The scale judgment subunit is connected to the exact solution subunit and the heuristic solution subunit. It counts the number of tasks to be matched and the number of online drones, compares the number of tasks to be matched and the number of online drones with a scale threshold, and selects either the exact solution subunit or the heuristic solution subunit to perform the solution operation based on the comparison results. When the number of tasks to be matched and the number of online drones are both less than the scale threshold, the exact solution subunit uses an integer programming algorithm to solve the multi-objective optimization model. When the number of tasks to be matched or the number of online drones is greater than or equal to the scale threshold, the heuristic solution subunit uses a heuristic search algorithm to solve the multi-objective optimization model.

[0063] The scale judgment subunit selects an appropriate solution algorithm based on the problem scale. The subunit first counts the number of currently unmatched tasks and the number of online drones, then compares these two numbers with a preset scale threshold. When both numbers are less than the threshold, the problem scale is small, and the exact solution subunit is selected to obtain the global optimum. When either number is greater than or equal to the threshold, the problem scale is large, and an exact solution takes too long. In this case, a heuristic solution subunit is selected to obtain an approximate optimum within an acceptable time. The scale threshold value needs to be determined based on computational resources and time requirements. A larger threshold allows for more scenarios using exact solutions and higher solution quality, but also increases computation time. The exact solution subunit uses integer programming algorithms to solve multi-objective optimization models, guaranteeing a global optimum and suitable for small-scale problems. Integer programming algorithms can employ mature algorithms such as branch and bound and cutting plane methods to ensure optimality of the solution. The heuristic solution subunit uses heuristic search algorithms to solve multi-objective optimization models, obtaining a high-quality approximate optimum in a shorter time and suitable for large-scale problems. Heuristic algorithms can employ intelligent optimization algorithms such as genetic algorithms, simulated annealing algorithms, and particle swarm optimization algorithms, or traditional heuristic methods such as greedy algorithms and local search. The specific choice of algorithm depends on the characteristics of the problem and the requirements for solving it.

[0064] The system also includes a scrolling window rematching module. This module is connected to the task urgency assessment module and the UAV suitability calculation module, and includes a status monitoring subunit, a trigger judgment subunit, and an incremental calculation subunit. The status monitoring subunit, connected to the trigger judgment subunit, monitors task addition events, task cancellation events, and task attribute change events on the task side, as well as UAV malfunction events, UAV battery fluctuation events, and UAV payload change events on the UAV side, generating a status change notification signal. The trigger judgment subunit, connected to the incremental calculation subunit, generates a rematching trigger signal based on the status change notification signal and the rematching trigger rules. The incremental calculation subunit determines the task set and UAV set affected by the status change based on the rematching trigger signal, performs an update operation on the multi-dimensional spatiotemporal coupling urgency and suitability values ​​for the affected task set and UAV set, and outputs the update result to the online matching optimization decision module.

[0065] The status monitoring subunit monitors various status change events on both the task side and the UAV side in real time, including task additions, cancellations, attribute changes, and UAV malfunctions, sudden power fluctuations, and payload changes. When a status change is detected, a status change notification signal is generated and sent to the trigger judgment subunit. The frequency of status monitoring can be set according to the system's real-time requirements; a higher frequency results in a faster response but also greater computational overhead. The trigger judgment subunit determines whether to initiate the rematch process based on preset rematch trigger rules. Trigger rules can be set to trigger directly for specific types of status changes or to trigger when status changes accumulate to a certain level, avoiding frequent rematches that could cause system oscillations. When the trigger conditions are met, a rematch trigger signal is generated and sent to the incremental calculation subunit. The setting of trigger rules needs to balance the timeliness of matching and the stability of the system; too frequent triggering can lead to system oscillations, while too sparse triggering can prevent timely responses to status changes. The incremental calculation subunit performs incremental calculations for rematching, first identifying the tasks and UAVs affected by status changes, and then only updating the urgency and suitability of the affected parts, rather than recalculating everything, thereby improving the response speed of rematching. After the update is complete, the results are output to the online matching optimization decision module for re-solution. Determining the scope of incremental calculations requires considering the propagation range of coupling effects to ensure that the affected parts are covered, while minimizing unnecessary calculations.

[0066] The system also includes an execution feedback and parameter tuning module. This module is connected to the task urgency assessment module and the UAV adaptability calculation module, and includes an execution data acquisition subunit, a deviation calculation subunit, and a parameter adaptive adjustment subunit. The execution data acquisition subunit is connected to the deviation calculation subunit and collects actual energy consumption data, actual completion time data, and actual completion quality data during the actual execution of each task. The deviation calculation subunit is connected to the parameter adaptive adjustment subunit and calculates the energy consumption deviation between the actual energy consumption data and the estimated energy consumption data, the time deviation between the actual completion time data and the estimated completion time data, and the quality deviation between the actual completion quality data and the estimated completion quality data. The parameter adaptive adjustment subunit generates urgency model parameter adjustment signals and adaptability model parameter adjustment signals based on the energy consumption deviation values, the time deviation values, and the quality deviation values. The urgency model parameter adjustment signals are output to the task urgency assessment module, and the adaptability model parameter adjustment signals are output to the UAV adaptability calculation module.

[0067] The execution data acquisition subunit collects various operational data during the actual execution of the task, including actual energy consumption, actual completion time, and actual completion quality. This data reflects the actual situation of task execution and serves as the basis for evaluating the accuracy of model predictions. The frequency and accuracy of data acquisition need to be determined based on the requirements of parameter tuning; the more accurate the data, the better the parameter tuning effect, but the higher the acquisition cost. The deviation calculation subunit compares the actual execution data with the model's predicted values ​​and calculates the deviation value. The energy consumption deviation value reflects the accuracy of the energy consumption prediction, the time deviation value reflects the accuracy of the time prediction, and the quality deviation value reflects the accuracy of the quality prediction. The larger the deviation, the less accurate the model prediction, and the greater the adjustment required. Deviation can be calculated using absolute deviation, relative deviation, or mean square deviation, etc. The specific form to be chosen depends on the requirements of the parameter tuning algorithm. The parameter adaptive adjustment subunit calculates the adjustment amount of the model parameters based on the deviation value and generates a parameter adjustment signal. The principle of adjustment is that the model parameters of the dimension with the larger deviation should be adjusted to make subsequent predictions more accurate. Parameter adjustment signals are sent to the task urgency assessment module and the UAV adaptability calculation module to adjust the corresponding model parameters. The parameter adjustment algorithm can employ gradient descent, least squares, or other adaptive algorithms to ensure that the parameters converge smoothly to their optimal values, avoiding oscillations. Through this closed-loop feedback mechanism, the model parameters can be continuously optimized as the system runs, gradually adapting to the actual scenario.

[0068] The system also includes a spatiotemporal constraint verification module. This module is connected to the task urgency assessment module and the online matching optimization decision module. It acquires time window data and spatial region data for each task to be matched, calculates the degree of spatiotemporal conflict between tasks based on the time window data and spatial region data, generates a spatiotemporal constraint matrix, and outputs the spatiotemporal constraint matrix to the online matching optimization decision module. The online matching optimization decision module also uses the spatiotemporal constraint matrix to solve the multi-objective optimization model.

[0069] The spatiotemporal constraint verification module pre-verifies and quantifies the spatiotemporal constraint relationships between tasks before matching optimization. The module first acquires the time window data and spatial region data for each task. The time window data defines the allowed time range for task execution, and the spatial region data defines the spatial range for task execution. Then, the module calculates the degree of spatiotemporal conflict between each pair of tasks; the greater the overlap of time windows and the closer the spatial regions, the higher the conflict degree. The calculation results of the conflict degree are stored in the form of a spatiotemporal constraint matrix, where each element corresponds to the conflict degree of a pair of tasks. After the spatiotemporal constraint matrix is ​​output to the online matching optimization decision module, the optimization process incorporates the constraint matrix to avoid assigning tasks with severe spatiotemporal conflicts to the same UAV, ensuring the executability of the matching scheme. The calculation of the spatiotemporal conflict degree can be determined based on a combination of time overlap rate and spatial distance; the greater the time overlap and the closer the spatial distance, the higher the conflict degree. The specific calculation method needs to be determined according to the spatiotemporal characteristics of the tasks. The introduction of the spatiotemporal constraint matrix allows matching optimization to consider not only suitability and benefits but also the constraint relationships between tasks, improving the feasibility and execution efficiency of the matching scheme.

[0070] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints, characterized in that, The system includes a task urgency assessment module, a UAV suitability calculation module, and an online matching optimization decision module. The task urgency assessment module is connected to the UAV suitability calculation module and is used to acquire attribute data of the set of tasks to be matched and calculate the multi-dimensional spatiotemporal coupling urgency of each task. The UAV suitability calculation module is connected to the online matching optimization decision module and is used to acquire the status data of the online UAV set and calculate the suitability value between each UAV and each task to be matched based on the multi-dimensional spatiotemporal coupling urgency. The online matching optimization decision module is used to construct a multi-objective optimization model based on the suitability values ​​and solve the multi-objective optimization model to obtain a task matching scheme.

2. The online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints according to claim 1, characterized in that, The task urgency assessment module includes a task data acquisition unit, an urgency calculation unit, and a coupled calculation unit. The task data acquisition unit is connected to the urgency calculation unit and is used to collect deadline data, coverage area data, resource requirement data, task type data, and geographic location coordinate data for each task to be matched. The urgency calculation unit is connected to the coupled calculation unit and is used to generate a time urgency component based on the deadline data, a spatial urgency component based on the coverage area data, and a resource urgency component based on the resource requirement data. The coupling calculation unit is used to calculate the coupling effect strength between each task to be matched based on the task type data and the geographic location coordinate data, and to generate the multidimensional spatiotemporal coupling urgency based on the time urgency component, the spatial urgency component and the resource urgency component.

3. The online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints according to claim 1, characterized in that, The drone adaptability calculation module includes a drone status acquisition unit, a single-factor evaluation unit, and a synergistic effect calculation unit; the drone status acquisition unit is connected to the single-factor evaluation unit and is used to collect the current location data, remaining power data, payload configuration data, total battery capacity data, and health status data of each online drone; The single-factor evaluation unit is connected to the synergistic effect calculation unit and is used to generate a distance reachability index based on the current location data, a power adequacy index based on the remaining power data and the total battery capacity data, and a load matching index based on the load configuration data. The synergy effect calculation unit is used to generate a synergy gain value based on the health status data and the assigned task information, and to generate the fit value based on the distance reachability index, the power adequacy index and the load matching index.

4. The online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints according to claim 1, characterized in that, The online matching optimization decision module includes an objective function construction unit, a constraint processing unit, and an optimization solution unit. The objective function construction unit is connected to the constraint processing unit and is used to construct a matching benefit term based on the fitness degree value, a non-matching penalty term based on the multi-dimensional spatiotemporal coupling urgency, and a load balancing penalty term based on the load distribution data of each UAV. It then generates an objective function based on the matching benefit term, the non-matching penalty term, and the load balancing penalty term. The constraint processing unit is connected to the optimization solution unit and is used to generate UAV task allocation constraints, task execution capability constraints, and time window constraints. The optimization solution unit is used to solve the multi-objective optimization model based on the objective function, the UAV task allocation constraints, the task execution capability constraints, and the time window constraints, and output the task matching scheme.

5. The online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints according to claim 2, characterized in that, The coupling calculation unit includes an iterative control subunit and a convergence judgment subunit. The iterative control subunit is connected to the convergence judgment subunit and is used to sequentially read the urgency data and surrounding task information of each task to be matched according to a preset iteration order, calculate the coupling correction urgency of each task to be matched, and write the coupling correction urgency into the storage area. The convergence judgment subunit is used to read the urgency data of each task to be matched in two adjacent iterations, calculate the sum of the urgency differences between the two iterations, and generate an iteration termination signal when the sum of the urgency differences is less than the convergence threshold and output it to the iterative control subunit.

6. The online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints according to claim 3, characterized in that, The single-factor evaluation unit further includes a weight adaptive adjustment subunit; the weight adaptive adjustment subunit is connected to the synergy effect calculation unit, and is used to acquire the current system's task quantity distribution data, task type distribution data, drone quantity distribution data, and drone status distribution data, generate distance dimension weight coefficients, battery dimension weight coefficients, and payload dimension weight coefficients based on the task quantity distribution data, task type distribution data, drone quantity distribution data, and drone status distribution data, and output the distance dimension weight coefficients, battery dimension weight coefficients, and payload dimension weight coefficients to the synergy effect calculation unit.

7. The online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints according to claim 4, characterized in that, The optimization solution unit includes a scale judgment subunit, an exact solution subunit, and a heuristic solution subunit. The scale judgment subunit is connected to the exact solution subunit and the heuristic solution subunit, and is used to count the number of tasks to be matched and the number of online drones, compare the number of tasks to be matched and the number of online drones with a scale threshold respectively, and select the exact solution subunit or the heuristic solution subunit to perform the solution operation based on the comparison result. The exact solution subunit is used to solve the multi-objective optimization model using an integer programming algorithm when the number of tasks to be matched and the number of online drones are both less than the scale threshold. The heuristic solution subunit is used to solve the multi-objective optimization model using a heuristic search algorithm when the number of tasks to be matched or the number of online drones is greater than or equal to the scale threshold.

8. The online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints according to claim 1, characterized in that, It also includes a scrolling window rematching module; the scrolling window rematching module is connected to the task urgency assessment module and the UAV adaptability calculation module, and includes a status monitoring subunit, a trigger judgment subunit and an incremental calculation subunit; The status monitoring subunit is connected to the trigger judgment subunit and is used to monitor task addition events, task cancellation events, and task attribute change events on the task side, as well as drone malfunction events, drone battery mutation events, and drone payload change events on the drone side, and generate status change notification signals. The trigger judgment subunit is connected to the incremental calculation subunit and is used to generate rematch trigger signals based on the status change notification signals and rematch trigger rules. The incremental calculation subunit is used to determine the task set and drone set affected by the status change based on the rematch trigger signals, perform update operations on the multidimensional spatiotemporal coupling urgency and the fit degree values ​​on the task set and drone set affected by the status change, and output the update results to the online matching optimization decision module.

9. The online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints according to claim 1, characterized in that, It also includes an execution feedback and parameter tuning module; the execution feedback and parameter tuning module is connected to the task urgency assessment module and the UAV adaptability calculation module, and includes an execution data acquisition subunit, a deviation calculation subunit and a parameter adaptive adjustment subunit; the execution data acquisition subunit is connected to the deviation calculation subunit and is used to collect actual energy consumption data, actual completion time data and actual completion quality data during the actual execution of each task. The deviation calculation subunit is connected to the parameter adaptive adjustment subunit and is used to calculate the energy consumption deviation between the actual energy consumption data and the estimated energy consumption data, the time deviation between the actual completion time data and the estimated completion time data, and the quality deviation between the actual completion quality data and the estimated completion quality data. The parameter adaptive adjustment subunit is used to generate urgency model parameter adjustment signals and adaptability model parameter adjustment signals based on the energy consumption deviation values, the time deviation values, and the quality deviation values. The urgency model parameter adjustment signals are output to the task urgency assessment module, and the adaptability model parameter adjustment signals are output to the UAV adaptability calculation module.

10. The online UAV mission matching system based on multidimensional evaluation and spatiotemporal constraints according to claim 1, characterized in that, It also includes a spatiotemporal constraint verification module; the spatiotemporal constraint verification module is connected to the task urgency assessment module and the online matching optimization decision module, and is used to acquire time window data and spatial region data of each task to be matched, calculate the degree of spatiotemporal conflict between tasks based on the time window data and the spatial region data, generate a spatiotemporal constraint matrix, and output the spatiotemporal constraint matrix to the online matching optimization decision module; the online matching optimization decision module is also used to solve the multi-objective optimization model in combination with the spatiotemporal constraint matrix.

Images