UAV and USV network resource allocation system and method based on improved whale algorithm

Abstract

The invention discloses a UAV and USV network resource allocation system and method based on an improved whale algorithm. A constraint condition construction module is used for taking a correlation coefficient between a USV and a UAV, a service duration of providing service for the UAV by the UAV and a processor calculation frequency of the USV as constraint conditions; the optimization problem construction module is used for constructing optimization problems of minimizing USV task processing energy consumption and maximizing USV data processing amount; and the resource allocation module is used for iteratively solving the optimization problem by using the improved whale algorithm to obtain resource allocation between the UAV and the USV network. According to the method, a traditional whale optimization algorithm is fused with a novel coding mode and crossover mutation operation, the problems of poor feasible solution performance, poor convergence and high time complexity in existing resource allocation are solved, and the QoS requirement of intensive tasks is met.

Description

Technical Field

[0001] The present invention relates to the technical field of UAV and USV collaborative network resource allocation, and in particular to a UAV and USV network resource allocation system and method based on an improved whale algorithm. Background Art

[0002] In recent years, unmanned surface vehicles (USVs) have achieved a series of landmark results in environmental monitoring, water search and rescue, and space resource detection by virtue of their technical advantages such as low cost, small size, high flexibility, and strong operability. At present, inland USVs integrate advanced sensor communication systems, computing systems, and sensor systems, and have certain autonomous perception and intelligent decision-making capabilities. They are expected to improve the informatization level of inland waters and are widely regarded as one of the key technologies of the new generation of shipping systems. However, USVs are limited by limited computing resources and battery capacity, and cannot effectively guarantee computing-intensive tasks such as intelligent navigation and path planning, and often cannot meet the quality of service (QoS) requirements. In addition, the inland shipping environment is complex and lacks telecommunications infrastructure. How to effectively guarantee the QoS requirements of USV computing tasks is an urgent problem to be solved.

[0003] In this context, the emergence of digital twin (DT)-driven UAV and USV networks provides a feasible solution to this technical problem. Unmanned Ariel Vehicle (UAV) can provide flexible data offloading services for USV, and DT can dynamically allocate computing and communication resources according to the operating status of UAV and USV. However, existing resource allocation methods generally have problems such as poor convergence and high time complexity, which inevitably lead to high energy consumption of USV and serious waste of computing resources.

[0004] Therefore, it is urgent to propose a UAV and USV network resource allocation method driven by digital twins, which can effectively meet the special requirements of the inland waterway shipping environment, that is, to ensure the QoS of processing computationally intensive tasks under the premise of meeting the constraints of computing resources and energy consumption; at the same time, to solve the problems of slow convergence, high computational complexity, and low performance in the resource allocation process. Finally, a low-complexity and high-efficiency UAV and USV network resource allocation mechanism is realized. Summary of the invention

[0005] The purpose of the present invention is to provide a UAV and USV network resource allocation system and method based on an improved whale algorithm. The present invention uses a data twin-driven USV and UAV collaborative network to dynamically allocate computing and communication resources according to the operating status of the UAV and USV, and uses the improved whale algorithm to solve the problems of poor feasible solution performance, poor convergence, and high time complexity in the existing resource allocation methods, so as to obtain a low-complexity UAV and USV network resource allocation method based on the improved whale optimization algorithm.

[0006] To achieve this purpose, the present invention designs a UAV and USV network resource allocation system based on an improved whale algorithm, comprising:

[0007] The constraint condition construction module is used to use the correlation coefficient between the USV and the UAV, the service duration of the service provided by the UAV to the UAV, and the processor calculation frequency of the USV as constraint conditions;

[0008] The optimization problem construction module is used to construct the optimization problem of minimizing the USV task processing energy consumption and maximizing the USV data processing volume;

[0009] The resource allocation module is used to obtain an improved whale algorithm by defining auxiliary variable optimization coding and adding crossover mutation operations in the whale algorithm, and use the improved whale algorithm to iteratively solve the optimization problem to obtain the global optimal whale individual position, and allocate resources between the UAV and USV networks according to the global optimal whale individual position;

[0010] Among them, the individual position of the whale represents a feasible solution to the optimization problem.

[0011] The beneficial effects of the present invention are:

[0012] 1) The present invention integrates the traditional whale optimization algorithm with a new encoding method and cross-mutation operation to allocate UAV and USV network resources, effectively reducing the algorithm complexity and improving the operating efficiency while meeting the QoS requirements of USV computing-intensive tasks. In addition, the present invention can effectively reduce the energy consumption of UAV and USV and extend the service time.

[0013] 2) The present invention proposes an improved whale optimization algorithm encoding mechanism, which converts the traditional encoding into a matrix into the form of auxiliary variables, avoiding the complex 0-1 constraints in the matrix and improving the search efficiency. When dealing with high-dimensional and complex constrained problems, traditional heuristic algorithms such as genetic algorithms have a very large search space and the solution-finding process is prone to fall into local optimality. By converting to the form of auxiliary variables, the search space can be effectively reduced, ensuring that the algorithm can converge to the global optimal solution or high-quality approximate solution faster.

[0014] 3) In order to solve the problem of poor convergence of traditional whale optimization algorithm for non-continuous variable optimization problems, the present invention proposes a novel constraint analysis and cross-mutation operation that integrates cross-compilation operation, introduces randomness and diversity in the training process, enhances the search capability of the whale optimization algorithm, and effectively avoids the algorithm from converging to the local optimal solution too early, thereby improving the global search capability. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic diagram of information interaction between the UAV and USV network resource allocation system based on the improved whale algorithm of the present invention;

[0016] Figure 2 A diagram of a collaborative network system of USV and UAV driven by DT in an embodiment of the present invention;

[0017] Figure 3 It is a schematic diagram of the coding mechanism of the improved whale optimization algorithm of the present invention;

[0018] Figure 4 It is a schematic diagram of the crossover mutation mechanism of the improved whale optimization algorithm of the present invention;

[0019] Figure 5 Schematic diagram of data processing volume achieved by various methods under different noises in the embodiments of the present invention;

[0020] Figure 6 This is a schematic diagram of the energy consumption of USV processing corresponding to the EWOA method of the patent and each method in the embodiment of the present invention. DETAILED DESCRIPTION

[0021] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0022] Embodiment 1:

[0023] like Figures 1 to 6 A UAV and USV network resource allocation system based on an improved whale algorithm is shown, comprising:

[0024] The constraint condition construction module is used to use the correlation coefficient between the USV and the UAV, the service duration of the service provided by the UAV to the UAV, and the processor calculation frequency of the USV as constraint conditions;

[0025] The optimization problem construction module is used to construct the optimization problem of minimizing the USV task processing energy consumption and maximizing the USV data processing volume;

[0026] The resource allocation module is used to obtain an improved whale algorithm by defining auxiliary variable optimization coding and adding crossover mutation operations in the whale algorithm, and use the improved whale algorithm to iteratively solve the optimization problem to obtain the global optimal whale individual position, and allocate resources between the UAV and USV networks according to the global optimal whale individual position;

[0027] Among them, the individual position of the whale represents a feasible solution to the optimization problem.

[0028] In the above technical scheme, the specific method of obtaining the improved whale algorithm by defining auxiliary variables to optimize the coding and adding crossover mutation operations in the whale algorithm is as follows: the correlation coefficient matrix is ​​formed by the correlation coefficients between the USV and the UAV, and the coding method of the elements in the correlation coefficient matrix is ​​converted into auxiliary variables composed of the UAV numbers provided by the USV; and the crossover mutation operation method is used to randomly select the individual positions of the whales for crossover mutation operations, thereby obtaining the improved whale algorithm.

[0029] In the above technical solution, the constraint conditions and optimization problem construction include: taking the correlation coefficient between USV and UAV, the service duration of the service provided by UAV to UAV and the processor calculation frequency of USV as constraints, the construction method is:

[0030] The UAV and USV network is set to include M USVs and N UAVs, all driven by DT. In each fixed time slot, the USV can unload the collected data to a UAV for processing. The correlation coefficient between USV and UAV is defined as Each UAV provides services to the USVs that choose to access it through time division multiplexing, and defines the service duration β; the processor calculation frequency of the USV is defined as f;

[0031] The constraints for the correlation coefficient α between USV and UAV, the service duration β of UAV providing services to UAV, and the processor calculation frequency f of USV are Specifically:

[0032]

[0033]

[0034] Among them, α m,n Indicates whether the mth USV chooses to offload data to the nth UAV within a fixed time slot; α m,n =1 means that the mth USV chooses to unload data to the nth UAV in this time slot; otherwise, α m,n =0, are the sets of USVs and UAVs respectively; in each time slot, the i-th USV can only unload data to one n-th UAV; β m represents the service duration of the mth USV accessing the UAV, δ is the time slot length; f m Calculate the frequency for the processor of the mth USV, and The core voltage of the USV processor can be adjusted dynamically. Indicates the maximum processor frequency of the mth USV;

[0035] The specific method for constructing the optimization problem of minimizing USV task processing energy consumption and maximizing USV data processing volume is:

[0036]

[0037] in, is the processor power consumption and transmission power consumption of USV, is the sum of the power consumption of all USV processors and the transmission power consumption, is the data processing amount of the mth USV, η is the weight factor, η<0.

[0038] The above construction minimizes the USV task processing energy consumption and maximizes the USV data processing amount. The specific process of the optimization problem and constraints is as follows:

[0039] The USV and UAV network contains M USVs and N UAVs, all driven by DT. The USV and UAV set is defined as and In each fixed time slot, the USV can offload the collected data to a UAV for processing. The correlation coefficient between the USV and the UAV is defined as For any α m,n =1 means that the mth USV chooses to unload data to the nth UAV in this time slot; otherwise, α m,n =0. Because α m,n is a binary variable, so it needs to satisfy:

[0040]

[0041] At the same time, in each time slot, USV i can only offload data to one nth UAV, so it needs to satisfy

[0042]

[0043] Each UAV provides services to the USVs that choose to access it through time division multiplexing, defining the service duration where β mrepresents the service duration of the mth USV accessing the UAV, β m Needs to be satisfied

[0044]

[0045] Where δ represents the time slot length. At the same time, the time slot length allocated to the USV by any n-th UAV must satisfy

[0046]

[0047] Assume that in this time slot, the mth USV chooses to offload data to the nth UAV, and the amount of offloaded data is It can be expressed as

[0048]

[0049] Among them, B m,n represents the access bandwidth allocated by the nth UAV to the mth USV, h0 is the channel gain, is the transmission power of the mth USV, d m,n represents the distance between the mth USV and the nth UAV, σ 2 Represents the ambient noise power. UAV is equipped with a high-performance processor, so the data processing delay is negligible.

[0050] In any time slot, the USV processor calculates the frequency controlled by piezoelectric, so f m Need to meet

[0051]

[0052] in, Indicates the maximum processor frequency of the mth USV. The local data processing capacity of the mth USV It can be expressed as

[0053]

[0054] Where ε represents the amount of data that can be processed per processor cycle. Define the energy consumption of the mth USV task processing as is the power consumption of the USV processor and the transmission power consumption. Therefore, it can be expressed as

[0055]

[0056] in, Indicates the effective capacitance switch of the USV processor. In any time slot, the mth USV data processing volume can be expressed as

[0057]

[0058] In order to extend the USV service cycle and reduce data backlog, an optimization problem is constructed to minimize the USV task processing energy consumption and maximize the USV data processing volume, and the correlation coefficient α between USV and UAV, service time β and USV processor calculation frequency f are jointly optimized. The optimization problem can be constructed as:

[0059]

[0060] Among them, η<0 is the weight factor. DT can obtain high-quality feasible solutions by solving the optimization problem, which can be used to control USV and UAV.

[0061] In the above technical solution, the resource allocation module can set the initialization process, initialize the number of UAVs to N, the number of USVs to M, randomly determine the initial position of the population, and the initial number of iterations r EWOA is 0, controls coefficients a and p1, and sets the maximum number of iterations.

[0062] The initialization process lays the foundation for the entire optimization process. By setting the number of UAVs to N, the number of USVs to M, and randomly determining the initial position of the population, the algorithm is ensured to explore the solution space from a diverse starting point, avoiding the algorithm search from falling into a local optimum or premature convergence. The initial number of iterations is set to 0, indicating that the algorithm is in the starting state, and the setting of the control coefficient helps to balance the relationship between global search and local development, allowing the algorithm to effectively search in a wide range of solution spaces while focusing on potential areas. The maximum number of iterations is used as a termination condition to prevent the algorithm from falling into an infinite loop and obtain results within a limited time.

[0063] In the resource allocation module of the above technical solution, the improved whale optimization algorithm encoding method is: establish a conversion relationship between the correlation coefficient matrix A and the auxiliary variable θ, convert the correlation coefficient matrix A into a combination of UAV numbers provided by each USV to form an auxiliary variable θ, where: θ m represents the UAV number that provides service for the mth USV and is an integer in the range [1, N]. The association coefficient matrix A is composed of variables α m,n The M×N matrix composed of

[0064] The whale group coordinates are encoded through the improved whale optimization algorithm encoding mechanism, specifically:

[0065] Assuming there are I whales in the group, the coordinates of the whale group are expressed as:

[0066]

[0067] in, Denotes the rthEWOA The coordinates after iterations, Specifically expressed as:

[0068]

[0069] Among them, the coordinates of each individual whale are composed of the feasible auxiliary variables θ of M USVs, the service time β and the processor calculation frequency f. Indicates the rth EWOA A set of feasible solutions corresponding to the mth USV in whale individual i after iterations.

[0070] Because the correlation coefficient matrix A is a matrix consisting of 0-1 variables α m,n The M×N matrix consists of a row with only one non-zero element. Traditional heuristic algorithms such as genetic algorithms are inefficient. The above solution converts the correlation coefficient matrix A into an auxiliary variable Among them, θ m It represents the UAV number that provides service for the mth USV and is an integer in the range of [1, N]. This form not only simplifies the representation, reduces the encoding dimension, avoids the complex 0-1 constraints in the matrix, but also improves the search efficiency. When dealing with high-dimensional and complex constrained problems, traditional heuristic algorithms such as genetic algorithms have a very large search space and are prone to fall into local optimal solutions. By converting to θ, the search space is greatly reduced, allowing the algorithm to find the global optimal solution or high-quality approximate solution faster.

[0071] Wherein, A represents an M×N matrix composed of correlation coefficients α, and α represents a general term for the correlation coefficient between USV and UAV.

[0072] In the above technical solution, the specific method of iteratively solving the optimization problem using the improved whale algorithm to obtain the best individual whale position of the current iteration is:

[0073] Calculate the fitness of each individual whale position, select the one with the largest fitness as the best individual whale position for the current iteration, and the rth position of whale i EWOA The fitness function of the coordinates after iterations for:

[0074]

[0075] in, is the sum of the power consumption of all USV processors and the transmission power consumption, is the power consumption and transmission power consumption of the mth USV processor; is the mth USV data processing amount, is the set of USVs in the USV and UAV network, η is the weight factor, η<0, Indicates the rth EWOA A set of feasible solutions corresponding to the mth USV in whale individual i after iterations.

[0076] Fitness evaluation analyzes the fitness value of each individual whale position through a specific fitness function to achieve the purpose of dynamically evaluating the quality of the solution. The core goals of the optimization problem: minimizing energy consumption and maximizing data processing volume are directly taken into consideration to ensure the consistency of the search direction; by introducing weight factors, the priority between the two goals can be flexibly adjusted in different application scenarios; selecting the whale individual with the largest fitness as the current best solution can ensure that the algorithm is evolving in a better direction at every step, increasing the possibility of finding the global optimal solution. It not only improves the efficiency and accuracy of the algorithm, but also enhances its ability to cope with complex optimization problems, especially in task allocation and resource management scenarios involving USV and UAV collaboration, effectively promoting the improvement of the overall performance of the system.

[0077] In the above technical solution, the specific method of using the improved whale algorithm to solve the optimization problem and obtain the global optimal whale individual position is:

[0078] The improved whale algorithm uses the fitness function to iteratively solve the optimization problem to obtain the best individual whale position in the iteration. According to the best individual whale position in the iteration and the control coefficients a and p1, the positions of other individual whales in the whale group except the current best individual whale are selected, and the improved whale algorithm is used again to solve the optimization problem until the maximum number of iterations is reached to obtain the global best individual whale position.

[0079] According to the best individual whale position under iteration and the control coefficients a and p1, the specific method for determining the positions of other individual whales in the whale group except the current best individual whale is as follows:

[0080] When a≥1, p1≥0.5, the whale's individual position is updated by surrounding the prey, and other individuals in the group move toward the current best whale individual position. The way to update the surrounding position can be expressed as:

[0081]

[0082] in, is the position of other individuals in the whale group in the next iteration, Indicates the rth EWOA In the iteration, the group of whales selects the individuals that are close to them, and in the process of surrounding the prey r1 and r2 are random numbers between 0 and 1; express The xth element in A1 represents the individual whale The position update direction, D x represents the update step size;

[0083] When a≥1, p1<0.5, during the whale hunting process, the whales in the group launch bubbles to surround the prey through spiral motion. Simulating the spiral motion to update the position can effectively increase the search ability of the individual, and other individuals in the group will The position is updated by spiral motion as the center. The position update method can be expressed as:

[0084]

[0085] in, is the position of other individuals in the whale group in the next iteration, b is the logarithmic spiral shape constant, and l is a random number ranging from -1 to 1;

[0086] When a<1, in order to prevent the group search from falling into the local optimum, the individuals in the group no longer move Instead, we randomly select an individual in the population to approach, and randomly select the rth individual. EWOA The position of an individual whale in iteration By updating the position formula of the surrounding prey, the distance between other whale individuals and random whale individuals is calculated. The surrounded position updates the position,

[0087] Among them, a decreases from 2 to 0 with the number of iterations, and p1 is a random number between 0 and 1.

[0088] This position update strategy effectively improves the performance of the optimization algorithm by dynamically adjusting the control coefficients a and p1, combining the three mechanisms of surrounding prey, spiral motion and random selection. In the early stage of iteration, when a is large, a≥1, and p1≥0.5, the method of surrounding prey is adopted to make the group approach the current optimal solution and accelerate the convergence speed; when p1<0.5, spiral motion is used to simulate hunting behavior, enhance local search capabilities, and avoid falling into local optimality too early. As the iteration proceeds, a gradually decreases to 0. When a<1, the group no longer relies solely on the current optimal solution, but randomly selects individuals as reference points to prevent the search process from converging prematurely, increasing the ability of global exploration.

[0089] It not only improves the probability of the algorithm finding the global optimal solution, but also enhances its robustness and adaptability, making it particularly suitable for complex and changeable optimization problems. In the task allocation of USV and UAV collaboration, it can efficiently balance the two goals of minimizing energy consumption and maximizing data processing volume, ensuring that the unmanned system achieves optimal performance when performing tasks. At the same time, this method also ensures that the algorithm has good exploration capabilities in different environments, thereby improving the overall optimization efficiency.

[0090] In the above technical solution, the crossover mutation operation added to the improved whale algorithm is:

[0091] The probability of crossover mutation at each whale individual position is CR, and the whale individual positions with p2≤CR are crossover mutation operations performed, and the following two crossover mutation methods are randomly selected for crossover mutation:

[0092] Crossover mutation method a: Randomly select the auxiliary variable θ corresponding to the UAV number of the mth USV in the whale individual i m , select [1,N] not equal to θ m The integer of is the new position of whale individual i;

[0093] Crossover mutation method b: Randomly select two feasible resource allocation solutions corresponding to the m-th USV and the m′-th USV in whale individual i and Swap the two to get the new individual whale positions.

[0094] Crossover mutation enhances the exploration capability of the whale optimization algorithm by introducing randomness and diversity. These two methods effectively prevent the algorithm from converging to the local optimal solution too early and improve the global search capability. In this way, crossover mutation not only increases the position diversity of the population, but also promotes the combination of excellent features of different feasible solutions, making the optimization process more flexible and efficient. It helps to find a better USV and UAV collaboration configuration, significantly reducing the energy consumption of task processing while increasing the amount of data processing.

[0095] To enhance the global search capability, the probability of crossover mutation for each whale individual i is CR, that is, when p2≤CR, the whale is subjected to crossover mutation operation, and p2 is a random number between 0 and 1. The two crossover mutation operation methods adopted in this patent are as follows: Figure 3 As shown, one of the two crossover mutation operations is randomly selected to generate a new whale position. The first crossover mutation operation is to randomly select the UAV number θ in the whale individual i that corresponds to the UAV that provides service to the mth USV. m , select [1,N] not equal to θ m The second crossover mutation operation is to randomly select two groups of USV resource allocation feasible solutions in whale individual i. and Swap the two to get the new individual whale positions.

[0096] In the above technical solution, in the resource allocation module, the method for analyzing the violation of constraints of the USV resource allocation scheme corresponding to the individual whales and repairing the position of the individual whales that do not meet the optimization problem of the constraint conditions is:

[0097] The individual positions generated by the execution of encircling prey, searching for prey, and bubble attack updates may violate the constraints. Therefore, the non-continuous auxiliary variable θ is analyzed and the distance The nearest integer in the range [1, N] is used as the coordinate of the repaired whale i;

[0098] For the constraint problem Analyze and find out what is not satisfactory The processor calculation frequency corresponding to the individual whale can be repaired by the following operation

[0099]

[0100] After completing the cross-compilation operation, check the constraint problem and Analyze if there is Then like Then

[0101] Among them, θ m,i , β m,i 、f m,i They represent the auxiliary variables, service duration and processor calculation frequency of the i-th feasible solution of the m-th USV, Indicates the rth EWOA The feasible solution corresponding to the updated whale individual in the iteration, Indicates the rth EWOA The feasible solution corresponding to the updated whale individual constraint processing in the iteration, where Indicates the rth EWOA The processor calculation frequency of the ith feasible solution of the mth USV of the updated whale individual in the iteration, Indicates the rth EWOA The processor calculation frequency of the i-th feasible solution for the m-th USV after the updated whale individual constraints in the iteration, Indicates the rth EWOA The service time of the i-th feasible solution of the m-th USV of the updated whale in the iteration, Indicates the rth EWOA The service time of the i-th feasible solution of the m-th USV of the updated whale in the iteration, Indicates the rth EWOA The auxiliary variable of the ith feasible solution of the mth USV of the whale individual after update in the iteration, represents the maximum frequency of the processor of the mth USV, and δ is the time slot length.

[0102] The repair operation enhances the robustness of the algorithm and avoids exploring invalid areas in the knowledge space, thereby improving the search efficiency. Ultimately, through strict constraint processing, the quality of the obtained solution is guaranteed, significantly improving the reliability and performance of the unmanned system collaborative task.

[0103] In the above technical solution, the resource allocation module also includes a decoding process, which is the inverse process of the encoding mechanism. The auxiliary variable θ in the global optimal whale individual position is decoded into the correlation coefficient α between the USV and the UAV by utilizing the conversion relationship between the correlation coefficient matrix A and the auxiliary variable θ, and is combined with the service time β and the processor calculation frequency f of the USV to obtain the global optimal feasible solution.

[0104] The decoding process is the inverse operation of the encoding process. In the above technical solution, the auxiliary variable θ in the global optimal whale individual position found by the optimization algorithm is decoded into the correlation coefficient α between the USV and the UAV, and combined with the service time β between the USV and the UAV and the processor calculation frequency f of the USV to output the global optimal feasible solution. It ensures that the optimization results can be directly applied to the actual system configuration, and realizes the seamless conversion from abstract mathematical models to specific parameter settings. In this way, not only the practicality and operability of the solution are improved, but also it is ensured that the optimized parameter configuration can effectively reduce energy consumption and improve data processing efficiency, thereby significantly enhancing the collaborative performance and task execution capabilities of the UAV and USV.

[0105] The above scheme is iterated. If the number of iterations reaches the maximum number of iterations, the iterative training ends.

[0106] Output the correlation coefficient α between USV and UAV, service time β and processor calculation frequency f of USV, and use the correlation coefficient α between USV and UAV, service time β and processor calculation frequency f of USV to optimize the resource allocation of USV and UAV at the physical layer through wireless interaction and control, such as Figure 2 As shown in the figure, DT runs in the remote cloud service center. The DT layer and the physical layer can interact in real time through wireless communication links, so the DT layer can obtain the coordinates and related physical parameters of the USV and UAV in real time and update the status of the twin individuals. In addition, the DT layer can issue operating instructions to the USV and UAV and allocate resources according to the demand status of the twin individuals. In each time slot, each USV will collect and process inland river hydrological data.

[0107] In order to better demonstrate the effectiveness and practicality of the Enhanced Whales Optimization Algorithm (EWOA) model proposed in this invention, the EWOA algorithm will be performance tested using MATLAB software, and compared with other advanced algorithms, such as Genetic Algorithm (GA), Differential evolution algorithm (DE) and Fixed Service Duration algorithm (FSD), to verify the superiority of the proposed method. Assume that the network coverage is 150m×150m, and 10 USVs are randomly distributed. Each service cycle is divided into 500 time slots, and the length of the time slots is 0.2s. The performance comparison results of the proposed EWOA and other algorithms are shown in the figure. Figure 5 As shown, Figure 5 The data processing volume achieved by each method under different noise environments is described. As the noise power increases, the USV data processing volume achieved by each method gradually decreases. This is because the increase in noise power will significantly reduce the signal-to-noise ratio of the USV to UAV transmission link, resulting in a decrease in data transmission volume. Specifically, when the noise power is -80dBm, FSD, GA, and EWOA achieve approximately 4×10 5 bits, 9×10 5 bits, 11×10 5 bits and 1×10 6 bits. EWOA can achieve the maximum USV data processing capacity and significantly reduce the backlog of USV data queue.

[0108] like Figure 6 The corresponding USV processing energy consumption of each method is described. It can be seen that EWOA can significantly reduce the USV processing energy consumption (about 50%) compared with other methods.

[0109] Embodiment 2:

[0110] A UAV and USV network resource allocation method based on an improved whale algorithm comprises the following steps:

[0111] Step 1, taking the correlation coefficient between USV and UAV, the service duration of the service provided by UAV to UAV and the processor calculation frequency of USV as constraints;

[0112] Step 2: Construct an optimization problem to minimize the USV task processing energy consumption and maximize the USV data processing volume;

[0113] Step 3: In the whale algorithm, an improved whale algorithm is obtained by defining auxiliary variable optimization coding and adding crossover mutation operations. The improved whale algorithm is used to iteratively solve the optimization problem to obtain the global optimal whale individual position, and the resources between the UAV and USV networks are allocated according to the global optimal whale individual position.

[0114] Among them, the individual position of the whale represents a feasible solution to the optimization problem.

[0115] Step 3.1, initialize the number of UAVs, the number of USVs, the population position, the initial number of iterations and the control coefficient, and set the maximum number of iterations. Initialize the number of UAVs to N, the number of USVs to M, randomly determine the initial position of the population, and the initial number of iterations to r. EWOA is 0, the control coefficients are a and p1, and the maximum number of iterations is set.

[0116] Step 3.2: Encode the whale group coordinates after each iteration of the individual whale position through the improved whale optimization algorithm encoding mechanism.

[0117] Step 3.3: Calculate the best individual whale position of the current iteration through the fitness function.

[0118] Step 3.4: According to the best individual whale position of the current iteration and the control coefficient, different stages are divided to update the positions of other individual whales by encircling the prey, attacking with a bubble net, and searching for prey.

[0119] Step 3.5: According to the probability of crossover mutation of each whale individual, use random numbers to screen the whale individuals for crossover mutation, perform crossover mutation operations on the feasible solutions for different USV resource allocations corresponding to the selected whale individual positions, and obtain the whale individual positions after crossover mutation.

[0120] Step 3.6: Analyze the constraint violations of the USV resource allocation scheme corresponding to the individual whales, and repair the positions of the individual whales that do not meet the constraint optimization problem.

[0121] Step 3.7, repeat steps 3.3 to 3.6 for iterative training. When the number of iterations reaches the maximum number of iterations, the iterative training ends and the global optimal whale individual position is obtained;

[0122] Step 3.8: Decode the global best individual whale position into the correlation coefficient between USV and UAV, the service time and the processor calculation frequency of USV.

[0123] In step 1 and step 2, the construction of the above constraints and optimization problems ensures the clarity and operability of the problem by clarifying the optimization goal (minimizing the USV task processing energy consumption and maximizing the data processing volume) and constraints (correlation coefficient, service time, processor calculation frequency). This provides a clear direction and basis for the subsequent algorithm design, ensuring the effectiveness and feasibility of the optimization process, thereby improving the quality of the solution and the convergence speed of the algorithm.

[0124] In step 3.1, the number of UAVs, the number of USVs, the population position, the initial number of iterations and the control coefficient are initialized, and the maximum number of iterations is set to provide the necessary starting point and boundary for the algorithm to run. This ensures that the algorithm starts from a reasonable starting point and finds a satisfactory solution within a limited number of iterations, avoiding infinite loops and improving the efficiency and practicality of the algorithm.

[0125] In step 3.2, the coordinates of the whale group after each iteration are encoded by the improved whale optimization algorithm (EWOA), which simplifies the problem representation and reduces the search space. This improves the search efficiency, enables the algorithm to converge to a high-quality solution more quickly, and reduces the computational complexity, enhancing the scalability and applicability of the algorithm.

[0126] In step 3.3, the fitness of each whale position is calculated through the fitness function, and the best whale individual in the current iteration is selected to ensure that there are high-quality solutions in each generation of the population. This guides the algorithm to evolve towards the optimal solution, improves the quality of the solution and the convergence speed of the algorithm, and helps to quickly find the global optimal solution or a high-quality approximate solution.

[0127] In step 3.4, based on the best individual whale position and control coefficient of the current iteration, the positions of other individual whales are updated using the strategies of encircling prey, attacking with bubble nets, and searching for prey. This enhances the global and local search capabilities of the algorithm, balances exploration and development, and can both find new solutions globally and refine local optimal solutions, thus improving the robustness of the algorithm and the quality of the solution.

[0128] In step 3.5, based on the probability of crossover mutation of each whale individual, random numbers are used to screen the crossover mutation whale individuals, and diversity is introduced by exchanging the feasible solutions of whale individuals corresponding to different USVs of interest. This increases the diversity of the population, prevents the algorithm from falling into the local optimum, improves the global search ability of the algorithm, enhances the robustness and adaptability of the algorithm, and helps to discover more potential high-quality solutions.

[0129] In step 3.6, the positions of individual whales that do not meet the constraints are repaired to ensure that all solutions meet the constraints. This ensures that the solutions generated by the algorithm are feasible, improves the quality of the solution and the robustness of the algorithm, avoids the generation of invalid solutions, and ensures the practical application value of the optimization results.

[0130] In step 3.7, the fitness function, position update module, crossover mutation and position repair are calculated in sequence, and iterative training is performed. When the number of iterations reaches the maximum number of iterations, the iterative training ends and the correlation coefficient between the USV and the UAV, the service time and the processor calculation frequency of the USV are output. This systematic iterative training ensures the stability and reliability of the algorithm and improves the overall optimization effect.

[0131] In step 3.8, the decoding operation is the inverse operation of the encoding operation, which can decode the auxiliary variable θ in the global optimal whale individual position into the correlation coefficient α between the USV and the UAV, and combine it with the service time β between the USV and the UAV and the processor calculation frequency f of the USV to output the global optimal feasible solution. It ensures that the optimization results can be directly applied to the actual system configuration, improves the practicality and operability of the solution, and also ensures that the optimized parameter configuration can effectively reduce energy consumption and improve data processing efficiency, thereby significantly enhancing the collaborative performance and task execution capabilities of the unmanned system.

[0132] The above scheme solves the multi-objective optimization problem of UAV and USV network resource allocation through the improved whale optimization algorithm (EWOA), aiming to minimize the energy consumption of USV task processing and maximize the data processing volume.

[0133] The resource allocation method based on the improved whale algorithm is:

[0134] S1. Input the number of initialized UAVs N and USVs M.

[0135] S2. Initialize population position Initial number of iterations r EWOA =0 and control coefficient a=2, and set the maximum number of iterations;

[0136] S3. Judgment Is it true? If so, go to step 4; if not, go to S11;

[0137] S4, determine whether a≥1 is true, if true, go to S5, if not true, go to S6;

[0138] S5, judge p1, if p1 ≥ 0.5, update the individual position of the whale by surrounding the prey, if p1 < 0.5, update the individual position of the whale by attacking with a bubble net;

[0139] S6. Update make Update individual whale positions by searching for prey;

[0140] S7. For violations of individual whales for repair;

[0141] S8. Update p2 and perform crossover and mutation operations on whale individuals with p2 < CR;

[0142] S9. Repair individuals that violate ;

[0143] S10. Update r EWOA and a, and loop S3 to S10;

[0144] S11. Output θ, β, f.

[0145] Embodiment 3:

[0146] A computer program product includes a computer program which, when executed by a processor, implements the steps of the above method.

[0147] The content not described in detail in this specification belongs to the prior art well-known to those skilled in the art. Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0148] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the specified functions in Figure 1 one or more of the flows Figure 1 or blocks or combinations of blocks.

[0149] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implements the specified functions in Figure 1 one or more of the flows Figure 1 or blocks or combinations of blocks.

[0150] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0151] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit its protection scope. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that after reading the present invention, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the pending claims of the invention.

[0152] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field.

Claims

1. A UAV and USV network resource allocation system based on an improved whale algorithm, characterized in that: include: The constraint condition construction module is used to use the correlation coefficient between the USV and the UAV, the service duration of the service provided by the UAV to the UAV, and the processor calculation frequency of the USV as constraint conditions; The optimization problem construction module is used to construct the optimization problem of minimizing the USV task processing energy consumption and maximizing the USV data processing volume; The resource allocation module is used to obtain an improved whale algorithm by defining auxiliary variable optimization coding and adding crossover mutation operations in the whale algorithm, and use the improved whale algorithm to iteratively solve the optimization problem to obtain the global optimal whale individual position, and allocate resources between the UAV and USV networks according to the global optimal whale individual position; Among them, the individual position of the whale represents a feasible solution to the optimization problem.

2. The UAV and USV network resource allocation system based on the improved whale algorithm according to claim 1 is characterized in that: In the whale algorithm, the specific method of obtaining the improved whale algorithm is to define auxiliary variables to optimize the coding and add crossover mutation operations: the correlation coefficients between USVs and UAVs constitute a correlation coefficient matrix, and the coding method of the elements in the correlation coefficient matrix is ​​converted into auxiliary variables composed of the UAV numbers provided by the USV; and the crossover mutation operation method is used to randomly select the individual positions of whales for crossover mutation operations.

3. The UAV and USV network resource allocation system based on the improved whale algorithm according to claim 1 is characterized in that: In the constraint condition construction module and the optimization problem construction module, the correlation coefficient between USV and UAV, the service duration of the service provided by UAV to UAV and the processor calculation frequency of USV are used as constraints. The construction method is: The UAV and USV network is set to include M USVs and N UAVs, all driven by DT. In each fixed time slot, the USV can unload the collected data to a UAV for processing. The correlation coefficient between USV and UAV is defined as Each UAV provides services to the USVs that choose to access it through time division multiplexing, and defines the service duration β; the processor calculation frequency of the USV is defined as f; The constraints for the correlation coefficient α between USV and UAV, the service duration β of UAV providing services to UAV, and the processor calculation frequency f of USV are Specifically: Among them, α m,n Indicates whether the mth USV chooses to offload data to the nth UAV within a fixed time slot; α m,n =1 means that the mth USV chooses to unload data to the nth UAV in this time slot; otherwise, α m,n =0, are the sets of USVs and UAVs respectively; in each time slot, the i-th USV can only unload data to one n-th UAV; β m represents the service duration of the mth USV accessing the UAV, δ is the time slot length; f m Calculate the frequency for the processor of the mth USV, and The core voltage of the USV processor can be adjusted dynamically; Indicates the maximum processor frequency of the mth USV; The specific method for constructing the optimization problem of minimizing USV task processing energy consumption and maximizing USV data processing volume is: in, is the voltage and transmit power consumption of the processor in USV, is the sum of the power consumption of all USV processors and the transmission power consumption, is the data processing amount of the mth USV, and η is the weight factor.

4. The UAV and USV network resource allocation system based on the improved whale algorithm according to claim 1 is characterized in that: In the resource allocation module, the improved whale optimization algorithm encoding method is: establish the conversion relationship between the correlation coefficient matrix A and the auxiliary variable θ, convert the correlation coefficient matrix A into the UAV number combination provided by each USV to form the auxiliary variable θ, where, θ m represents the UAV number that provides service for the mth USV and is an integer in the range [1, N]. The association coefficient matrix A is composed of variables α m,n The M×N matrix composed of The whale group coordinates are encoded through the improved whale optimization algorithm encoding mechanism, specifically: Assuming there are I whales in the group, the coordinates of the whale group are expressed as: in, Represented as the rth EWOA The coordinates after iterations, Specifically expressed as: Among them, the coordinates of each individual whale are composed of the feasible auxiliary variables θ of M USVs, the service time β and the processor calculation frequency f. Indicates the rth EWOA A set of feasible solutions corresponding to the mth USV in whale individual i after iterations.

5. The UAV and USV network resource allocation system based on the improved whale algorithm according to claim 1 is characterized in that: The specific method of iteratively solving the optimization problem using the improved whale algorithm to obtain the best individual whale position in the current iteration is: Calculate the fitness of each individual whale position, select the one with the largest fitness as the best individual whale position for the current iteration, and the rth position of whale i EWOA The fitness function of the coordinates after iterations for: in, is the power consumption of the processor and the transmission power consumption in the mth USV, is the sum of the power consumption of all USV processors and the transmission power consumption, is the mth USV data processing amount, is the set of USVs in the USV and UAV network, η is the weight factor, Indicates the rth EWOA A set of feasible solutions corresponding to the mth USV in whale individual i after iterations.

6. The UAV and USV network resource allocation system based on the improved whale algorithm according to claim 5 is characterized in that: The specific method of using the improved whale algorithm to solve the optimization problem and obtain the global optimal whale individual position is: The improved whale algorithm uses the fitness function to iteratively solve the optimization problem to obtain the best individual whale position in the iteration. According to the best individual whale position in the iteration and the control coefficients a and p1, the positions of other individual whales in the whale group except the current best individual whale are selected, and the improved whale algorithm is used again to solve the optimization problem until the maximum number of iterations is reached to obtain the global best individual whale position. According to the best individual whale position under iteration and the control coefficients a and p1, the specific method for determining the positions of other individual whales in the whale group except the current best individual whale is as follows: When a≥1, p1≥0.5, the position of the individual whale is updated by surrounding the prey, and the other individuals in the group move towards the best whale individual. The way to update the surrounding position can be expressed as: in, is the position of other individuals in the whale group in the next iteration, Indicates the rth EWOA In the iteration, the group of whales selects the individuals that are close to them, and in the process of surrounding the prey r1 and r2 are random numbers between 0 and 1; express The xth element in A1 represents the individual whale The position update direction, D x represents the update step size; When a≥1, p1<0.5, other individuals in the group will The position is updated through spiral motion as the center. The position update method is expressed as: in, is the position of other individuals in the whale group in the next iteration, b is the logarithmic spiral shape constant, and l is a random number ranging from -1 to 1; When a<1, randomly select the rth EWOA The position of an individual whale in iteration By updating the position formula of the surrounding prey, the distance between other whale individuals and random whale individuals is calculated. The surrounding position updates the position, where 7. The UAV and USV network resource allocation system based on the improved whale algorithm according to claim 6 is characterized in that: In the improved whale algorithm, the added crossover mutation operation is: The probability of crossover mutation at each whale individual position is CR, and the whale individual positions with p2≤CR are crossover mutation operations performed, and the following two crossover mutation methods are randomly selected for crossover mutation: Crossover mutation method a: Randomly select the auxiliary variable θ corresponding to the UAV number of the mth USV in the whale individual i m , select [1,N] that is not equal to θ m The integer of is the new position of whale individual i; Crossover mutation method b: Randomly select two feasible resource allocation solutions corresponding to the m-th USV and the m′-th USV in whale individual i and Swap the two to get the new individual whale positions.

8. The UAV and USV network resource allocation system based on the improved whale algorithm according to claim 7 is characterized in that: In the resource allocation module, the constraint violation of the USV resource allocation scheme corresponding to the individual whale is analyzed, and the method for repairing the position of the individual whale that does not meet the optimization problem of the constraint condition is as follows: Analyze the non-continuous auxiliary variable θ and take the distance The nearest integer in the range [1, N] is used as the coordinate of the repaired whale i; For the constraint problem Analyze and find out what is not satisfactory The processor calculation frequency corresponding to the individual whale can be repaired by the following operation After completing the cross-compilation operation, check the constraint problem and Analyze if there is Then like Then Among them, θ m,i , β m,i 、f m,i They represent the auxiliary variables, service duration and processor calculation frequency of the i-th feasible solution of the m-th USV, Indicates the rth EWOA The feasible solution corresponding to the updated whale individual in the iteration, Indicates the rth EWOA The feasible solution corresponding to the updated whale individual constraint processing in the iteration, represents the maximum frequency of the processor of the mth USV, and δ is the time slot length.

9. A UAV and USV network resource allocation method based on an improved whale algorithm, characterized in that: The following steps are involved: The correlation coefficient between USV and UAV, the service duration of the service provided by UAV to UAV and the processor calculation frequency of USV are used as constraints; Construct an optimization problem to minimize the USV task processing energy consumption and maximize the USV data processing volume; In the whale algorithm, an improved whale algorithm is obtained by defining auxiliary variable optimization coding and adding crossover mutation operations. The improved whale algorithm is used to iteratively solve the optimization problem to obtain the global optimal whale individual position, and resources are allocated between the UAV and USV networks according to the global optimal whale individual position. Among them, the individual position of the whale represents a feasible solution to the optimization problem.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method described in claim 9 are implemented.

Images