WO2021035911A1 - 一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统 - Google Patents

一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统 Download PDF

Info

Publication number
WO2021035911A1
WO2021035911A1 PCT/CN2019/112603 CN2019112603W WO2021035911A1 WO 2021035911 A1 WO2021035911 A1 WO 2021035911A1 CN 2019112603 W CN2019112603 W CN 2019112603W WO 2021035911 A1 WO2021035911 A1 WO 2021035911A1
Authority
WO
WIPO (PCT)
Prior art keywords
genetic algorithm
path planning
data
probability
unmanned boat
Prior art date
Application number
PCT/CN2019/112603
Other languages
English (en)
French (fr)
Inventor
辛峻峰
张永波
李鹏昊
李世鑫
杨奉儒
Original Assignee
青岛蓝海未来海洋科技有限责任公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 青岛蓝海未来海洋科技有限责任公司 filed Critical 青岛蓝海未来海洋科技有限责任公司
Publication of WO2021035911A1 publication Critical patent/WO2021035911A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • G06Q10/04Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G06Q10/047Optimisation of routes or paths, e.g. travelling salesman problem
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01CMEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/20Instruments for performing navigational calculations
    • G01C21/203Specially adapted for sailing ships
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/12Computing arrangements based on biological models using genetic models
    • G06N3/126Evolutionary algorithms, e.g. genetic algorithms or genetic programming
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/40Business processes related to the transportation industry

Definitions

  • the present disclosure belongs to the technical field of unmanned boat path planning, and relates to an unmanned boat path planning method and system based on a linear variable parameter genetic algorithm driven by data forward and backward.
  • GA genetic algorithm
  • CGA traditional genetic algorithm
  • one or more embodiments of the present disclosure provide a method and system for unmanned boat path planning based on data forward and backward driving linear variable parameter genetic algorithm, which can drive linear variable forward and backward through data.
  • Parametric genetic algorithm or data back-driven linear variable parameter genetic algorithm adaptively adjust the crossover probability and mutation probability, effectively and quickly traverse all candidate solutions, find the optimal path in a broader field, and have a shorter optimal path , Faster convergence speed and better robustness.
  • a path planning method for an unmanned craft based on a linear variable parameter genetic algorithm driven by data forward and backward.
  • An unmanned craft path planning method based on data forward and backward driving linear variable parameter genetic algorithm includes:
  • the optimal trajectory of the UAV path planning is generated by using the data forward driving linear change parameter genetic algorithm or the data forward driving linear change parameter genetic algorithm.
  • the navigation data of the unmanned boat is the latitude and longitude values collected by the navigation system of the unmanned boat, and the specific steps of converting the navigation data into a coordinate system and drawing a coordinate map include:
  • the specific steps of generating the optimal trajectory of the UAV path planning by using the data forward driving linear change parameter genetic algorithm include:
  • the parameters include population size, maximum number of iterations, current number of iterations, crossover probability, and mutation probability;
  • the optimal individual is selected from the retention results of each iteration as the optimal solution of the genetic algorithm with forward-driven linear change parameters, otherwise the iteration is continued.
  • the adaptive crossover probability is calculated according to the product of the current iteration number and the difference between the final value of the crossover probability and the initial value divided by the sum of the maximum number of iterations and the initial value of the crossover probability;
  • the adaptive mutation probability is calculated according to the product of the current iteration number and the difference between the final value of the mutation probability and the initial value divided by the sum of the maximum iteration number and the initial value of the mutation probability.
  • the specific steps of using the genetic algorithm to drive the linearly variable parameters of the data back to generate the optimal trajectory of the UAV path planning include:
  • the parameters include population size, maximum number of iterations, current number of iterations, crossover probability, and mutation probability;
  • the optimal individual is selected from the retention results of each iteration as the optimal solution of the genetic algorithm with forward-driven linear change parameters, otherwise the iteration is continued.
  • the adaptive crossover probability is calculated according to the product of the difference between the maximum number of iterations and the current number of iterations and the final value of the crossover probability and the difference between the initial value divided by the sum of the maximum number of iterations and the initial value of the crossover probability;
  • the adaptive mutation probability is calculated according to the product of the current iteration number and the difference between the final value of the mutation probability and the initial value divided by the sum of the maximum iteration number and the initial value of the mutation probability.
  • the method also includes integrating the optimal trajectory of the unmanned boat path planning with ship log and status information to obtain the navigation trajectory of the unmanned boat and the distance and deflection angle from the target point to control the steering gear. Track offset correction and ship speed adjustment.
  • a computer-readable storage medium is provided.
  • a computer-readable storage medium in which a plurality of instructions are stored, and the instructions are suitable for being loaded by a processor of a terminal device and executing the unmanned boat path based on the data forward and reverse driving linear variable parameter genetic algorithm Planning method.
  • a terminal device is provided.
  • a terminal device which includes a processor and a computer-readable storage medium, the processor is used to implement each instruction; the computer-readable storage medium is used to store a plurality of instructions, the instructions are suitable for being loaded by the processor and executing the one An unmanned craft path planning method based on data forward and backward driving linear variable parameter genetic algorithm.
  • an unmanned boat path planning device based on data forward and reverse driving linear variable parameter genetic algorithm.
  • An unmanned boat path planning device based on data forward and backward driving linear variable parameter genetic algorithm based on the unmanned boat path planning method based on data forward and backward driving linear variable parameter genetic algorithm, including:
  • the data acquisition module is configured to receive navigation data, ship logs and status information of the unmanned boat;
  • the data processing module is configured to convert the navigation data into the coordinate system and draw a coordinate map, mark the target point and the real-time position of the unmanned boat in the coordinate map, and integrate the ship log and status information;
  • the path planning module is configured to generate the optimal trajectory of the UAV's path planning by adopting the genetic algorithm for data forward driving linear change parameters or the genetic algorithm for data forward driving linear change parameters.
  • the present disclosure provides a method and system for unmanned boat path planning based on data forward and reverse driving linear variable parameter genetic algorithm, which overcomes the disadvantages of traditional genetic algorithm such as slow convergence speed, low calculation efficiency, and premature convergence (local optimization).
  • traditional genetic algorithm such as slow convergence speed, low calculation efficiency, and premature convergence (local optimization).
  • the two improved algorithms can effectively avoid the local optimum, maintain a higher population diversity in the subsequent iteration process, reduce the average optimal distance, and have better robustness, especially This is when there are many planning points.
  • FIG. 1 is a flow chart of a path planning method for an unmanned boat based on a data forward-driven linear variable parameter genetic algorithm provided by one or more embodiments of the present disclosure
  • Figure 2 is a flow chart of a genetic algorithm provided by one or more embodiments of the present disclosure
  • FIG. 3 is an optimal path pair diagram under five types of points to be planned provided by one or more embodiments of the present disclosure
  • FIG. 4 is a diagram of the average path distance between the optimal path distance and the number of points to be planned according to one or more embodiments of the present disclosure
  • FIG. 5 is a variance diagram of the five optimal path distances and the optimal path distance of the number of points to be planned according to one or more embodiments of the present disclosure
  • FIG. 6 is a comparison diagram between the optimal path distance provided by one or more embodiments of the present disclosure and the Monte Carlo simulation quantity with five population sizes;
  • FIG. 7 is a comparison diagram of the optimal path distance and the average value of the overall size provided by one or more embodiments of the present disclosure.
  • FIG. 8 is a comparison diagram of the average value and variance value of the optimal path distance and the overall size provided by one or more embodiments of the present disclosure.
  • FIG. 9 is a graph of convergence curves of three algorithms under 50 planning points provided by one or more embodiments of the present disclosure.
  • FIG. 10 is an optimal path diagram of three algorithms under 50 planning points provided by one or more embodiments of the present disclosure.
  • Fig. 11 is an optimal trajectory diagram of four planning points during a USV test provided by one or more embodiments of the present disclosure.
  • each block in the flowchart or block diagram may represent a module, program segment, or part of code, and the module, program segment, or part of code may include one or more for implementing various embodiments. Executable instructions for the specified logic function.
  • the functions marked in the block may also occur in a different order from the order marked in the drawings. For example, two blocks shown in succession may actually be executed substantially in parallel, or they may sometimes be executed in the reverse order, depending on the functions involved.
  • each block in the flowchart and/or block diagram, and the combination of the blocks in the flowchart and/or block diagram can be implemented using a dedicated hardware-based system that performs the specified functions or operations. Or it can be implemented using a combination of dedicated hardware and computer instructions.
  • CGA Traditional genetic algorithm
  • CGA The traditional genetic algorithm (CGA) selects the method of real number encoding, and uses a string with the serial number of the visited city to represent each chromosome. Genetic parameters, such as population size, crossover and mutation probability, are usually defined empirically. After the optimization problem is determined, an initial population of candidate solutions with a certain scale is randomly generated. The fitness function is 1/len (len represents the relative path length of each chromosome), which is used to evaluate the fitness of each individual, and more suitable individuals will survive reproduction. Then, the algorithm uses iterative operations of crossover, mutation, and selection to increase the population size. If a certain standard is met or the maximum number of iterations is reached, the evolutionary process will be terminated.
  • crossover is mainly used to connect two parent chromosomes. These chromosomes are separated by a certain breakpoint and produce two offspring with a certain crossover probability (PC). Mutations are mainly used to exchange gene positions on two randomly selected mutation points on the chromosome, and the occurrence of mutations has a certain mutation probability (PM). It should be pointed out that crossover makes the chromosomes similar and helps the convergence of the population; while mutation increases the genetic diversity, so that the algorithm can further expand the population number under the condition of local optimization.
  • an unmanned craft path planning method based on data forward and backward driving linear variable parameter genetic algorithm includes:
  • An unmanned craft path planning method based on data forward and backward driving linear variable parameter genetic algorithm includes:
  • Step S1 Receive the navigation data, ship log and status information of the unmanned boat
  • Step S2 Convert the navigation data to the coordinate system and draw a coordinate map, and mark the target point and the real-time position of the unmanned boat in the coordinate map;
  • Step S3 Use the data forward-driven linear change parameter genetic algorithm or the data forward drive linear change parameter genetic algorithm to generate the optimal trajectory of the UAV path planning.
  • Step S4 Integrate the optimal trajectory of the UAV path planning with the ship log and status information to obtain the trajectory of the UAV as well as the distance and deflection angle from the target point, which is used to control the steering gear to correct the track offset And ship speed adjustment.
  • the navigation data of the unmanned boat is the latitude and longitude values collected by the unmanned boat navigation system.
  • step S2 of this embodiment the specific steps of converting the navigation data into a coordinate system and drawing a coordinate map include:
  • the concept of "change acceleration coefficient" A.Ratnaweera et al proposed improved.
  • a larger individual cognitive coefficient and a smaller social cognitive coefficient are used to allow the particles to search as much space as possible and avoid the algorithm from converging to a local optimum.
  • the coefficient of individual cognition decreases, while the coefficient of social cognition increases to ensure that the algorithm can successfully converge within the controllable range and avoid the situation where the algorithm cannot converge.
  • step S3 of this embodiment the genetic operator parameters of the genetic algorithm vary linearly with the number of iterations. According to the data-driven direction, improvements are divided into two categories, namely forward type (LCPGA) and reverse type (RLCPGA). Both types can help individuals traverse all candidate solutions and find the optimal solution in a wide range of fields.
  • LCPGA forward type
  • RLCPGA reverse type
  • Equation (1) and (2) represent updated mathematical correlation between P C and P M values:
  • the specific steps of generating the optimal trajectory of the UAV's path planning by using the genetic algorithm of data forward driving linear change parameters include:
  • Step S3-1 Perform parameter initialization.
  • the parameters include population size, maximum number of iterations, current iteration number, crossover probability, and mutation probability; set population size N, maximum number of iterations MAXGEN, current number of iterations k, and initial crossover probability P c , The initial mutation probability P m ;
  • Step S3-3 Calculate the fitness value and sort the initial population; the fitness function is defined as 1/len, where len represents the relative path length of each chromosome. Sort the initial population according to the calculated fitness value.
  • Step S3-4 Calculate the adaptive crossover probability and mutation probability. As the number of iterations increases, the crossover probability linearly decreases from 0.9 to 0.1, and the mutation probability linearly increases from 0.001 to 0.1;
  • the adaptive crossover probability is calculated according to the product of the current iteration number and the difference between the final value of the crossover probability and the initial value divided by the sum of the maximum number of iterations and the initial value of the crossover probability; according to the current iteration number and the difference between the final value of the mutation probability and the initial value
  • the product of the value is divided by the sum of the maximum number of iterations and the initial value of the mutation probability to calculate the adaptive mutation probability.
  • the following formula is used to specifically calculate the adaptive crossover probability and mutation probability:
  • Step S3-5 Perform adaptive crossover and mutation operations on the chromosomes according to the adaptive crossover probability and mutation probability; after calculating the obtained crossover probability and mutation probability, perform adaptive crossover and mutation operations on the chromosome.
  • the crossover operator uses the cyclic crossover operator
  • the mutation operator uses the gene swap mutation operator. Calculate the fitness value of the new offspring, re-order according to the value, and obtain the new population to enter the termination condition judgment.
  • Step S3-6 Determine whether the iteration termination condition is satisfied; the iteration termination condition is set to meet the requirements of a certain working condition or the number of iterations reaches the maximum. If the termination condition is not met, the number of iterations is increased by one, and step S3-3 is transferred; if it is met, step S3-7 is transferred.
  • Step S3-7 Select the optimal individual from the retention results of each iteration as the optimal solution of the forward-driven linear change parameter genetic algorithm and output, and the whole algorithm ends.
  • the initial value P C and P M are respectively provided 0.1 to 0.9 in the range of 0.001 to 0.1 .
  • P C increased from 0.1 linear 0.9
  • P M increases linearly from 0.001 to 0.1.
  • a smaller and a larger P C P M at the start of the initial iteration will help to maintain the diversity of population, and to help individuals through all candidates and solutions.
  • a larger and a smaller P C P M will help individuals with faster convergence clustered around the global optimum.
  • Formula (3) shows that, in accordance with the iterative sequence number, P C values may be adaptively optimized.
  • P C In order to protect the diversity of randomly generated groups, improve the ability of global search, P C is small in the initial iteration, when all individuals tend to be consistent with the increasing iterations, P C will increase the ability of local search Increased accordingly.
  • step S3 of this embodiment the specific steps of generating the optimal trajectory of the UAV's path planning by using the data back-driving linear change parameter genetic algorithm include:
  • the fitness function is defined as 1/len, where len represents the relative path length of each chromosome. Sort the initial population according to the calculated fitness value.
  • P C values may be adaptively optimized.
  • P C is small in the initial iteration, when all individuals tend to be consistent with the increasing iterations, P C will increase the ability of local search Increased accordingly.
  • the adaptive mutation probability is calculated according to the product of the current iteration number and the difference between the final value of the mutation probability and the initial value divided by the sum of the maximum iteration number and the initial value of the mutation probability.
  • subscript e and s respectively represent the final and initial values of P C and P M
  • k is the current iteration number
  • MAXGEN represents the maximum number of iterations.
  • C5. After calculating the obtained crossover probability and mutation probability, perform adaptive crossover and mutation operations on the chromosomes. In order to ensure that the gene code in each chromosome appears and only appears once, the crossover operator uses the cyclic crossover operator, and the mutation operator uses the gene swap mutation operator. Calculate the fitness value of the new offspring, re-order according to the value, and obtain the new population to enter the termination condition judgment.
  • the iteration termination condition is set to meet the requirements of a certain working condition or the number of iterations reaches the maximum. If the termination condition is not met, the number of iterations is increased by one and go to C3; if it is met, go to C7.
  • the optimal individual is selected from the retention results of each iteration as the optimal solution of the back-driving linear change parameter genetic algorithm and output, and the whole algorithm ends.
  • the planning point (P) considers five different population size programs, and the specific population number programs are: 10, 20, 30, 40, and 50.
  • the overall size S and the maximum number of iterations of each algorithm are set to 100 and 1000, respectively.
  • 100 Monte Carlo simulation experiments were carried out for each algorithm and working condition.
  • Figure 3 shows the distribution of the number of Monte Carlo simulations (N) and the best path distance (D) for different algorithms.
  • the blue symbols of the numbers in the left, middle, and right columns represent the data calculated by CGA, LCPGA, and RLCPGA, respectively.
  • the average value of 100 simulations is indicated by a red line in each graph. The result shows that with the increase of the P value, the distribution range of the symbols on the average line becomes larger and larger, and the ordinate value of the straight line becomes larger and larger.
  • p is set to 10
  • the average line of the two improved algorithms is significantly lower than the average line of the CGA algorithm.
  • the average value of the optimal path distance obtained by different algorithms is given And variance (V), to compare the path planning capabilities of different algorithms from a quantitative perspective.
  • V And variance
  • the path planning performance of two improved algorithms under 5 population sizes (20, 30, 60, 80, and 100) will be considered.
  • the number of planning points and the maximum number of iterations of each algorithm are set to 50 and 1000, respectively, and 100 Monte Carlo simulations are performed for each algorithm.
  • the optimal path distance (D) for different populations (S) under different simulation numbers (N) is shown for 100 Monte Carlo experiments, and the simulation data are scattered around the average line.
  • the results show that for five different population sizes, the simulated data in the improved algorithm is closer to the average line than the data in CGA, and the average line of the LCPGA and RLCPGA algorithms is significantly lower than that of the CGA.
  • the difference between the degree of data dispersion and the average value of the optimal path of the two improved algorithms is relatively small, which cannot be directly seen from Figure 5. Therefore, the quantitative results must be presented based on the average value D and the variance V, as shown in Figure 7 and Shown in Figure 8.
  • the CGA variances of various cluster sizes are maintained at the highest level above 0.5m 2 and decrease with the increase of S.
  • the variance curve of the improved algorithm has a similar oscillation trend, but the variance value is always less than 0.1.
  • Figure 10 shows the trajectories of 50 planning points designed by each algorithm.
  • the abscissa and ordinate respectively represent the latitude and longitude of each planned point, the starting point is enclosed in a red rectangle, and the arrow represents the direction of the planned path. It can be found that in Figure 10, only the trajectory generated by CGA has obvious path crossing phenomenon, which is why the path distance generated by the same conditions is longer than other algorithms. But at the same time, the advantages of the improved algorithm are more obvious in terms of effectively avoiding path crossings and simplifying path shapes.
  • a computer-readable storage medium is provided.
  • a computer-readable storage medium in which a plurality of instructions are stored, and the instructions are suitable for being loaded by a processor of a terminal device and executing the unmanned boat path based on the data forward and reverse driving linear variable parameter genetic algorithm Planning method.
  • a terminal device is provided.
  • a terminal device which includes a processor and a computer-readable storage medium, the processor is used to implement each instruction; the computer-readable storage medium is used to store a plurality of instructions, the instructions are suitable for being loaded by the processor and executing the one An unmanned craft path planning method based on data forward and backward driving linear variable parameter genetic algorithm.
  • the device executes the method or process described in the various embodiments of the present disclosure.
  • the computer program product may include a computer-readable storage medium loaded with computer-readable program instructions for executing various aspects of the present disclosure.
  • the computer-readable storage medium may be a tangible device that can hold and store instructions used by the instruction execution device.
  • the computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing.
  • Non-exhaustive list of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) Or flash memory), static random access memory (SRAM), portable compact disk read-only memory (CD-ROM), digital versatile disk (DVD), memory stick, floppy disk, mechanical encoding device, such as a printer with instructions stored thereon
  • RAM random access memory
  • ROM read-only memory
  • EPROM erasable programmable read-only memory
  • flash memory flash memory
  • SRAM static random access memory
  • CD-ROM compact disk read-only memory
  • DVD digital versatile disk
  • memory stick floppy disk
  • mechanical encoding device such as a printer with instructions stored thereon
  • the computer-readable storage medium used here is not interpreted as the instantaneous signal itself, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (for example, light pulses through fiber optic cables), or through wires Transmission of electrical signals.
  • the computer-readable program instructions described in the present disclosure can be downloaded from a computer-readable storage medium to various computing/processing devices, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and/or a wireless network.
  • the network may include copper transmission cables, optical fiber transmission, wireless transmission, routers, firewalls, switches, gateway computers, and/or edge servers.
  • the network adapter card or network interface in each computing/processing device receives computer-readable program instructions from the network, and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing/processing device .
  • the computer program instructions used to perform the operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or in one or more programming languages
  • Source code or object code written in any combination of, the programming language includes object-oriented programming languages—such as C++, etc., and conventional procedural programming languages—such as "C" language or similar programming languages.
  • Computer-readable program instructions can be executed entirely on the user's computer, partly on the user's computer, executed as a stand-alone software package, partly on the user's computer and partly executed on a remote computer, or entirely on the remote computer or server carried out.
  • the remote computer can be connected to the user's computer through any kind of network-including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (for example, using an Internet service provider to connect to the user's computer) connection).
  • LAN local area network
  • WAN wide area network
  • an electronic circuit such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), can be customized by using the status information of the computer-readable program instructions.
  • FPGA field programmable gate array
  • PDA programmable logic array
  • the computer-readable program instructions are executed to implement various aspects of the present disclosure.
  • an unmanned boat path planning device based on data forward and reverse driving linear variable parameter genetic algorithm.
  • An unmanned boat path planning device based on data forward and backward driving linear variable parameter genetic algorithm based on the unmanned boat path planning method based on data forward and backward driving linear variable parameter genetic algorithm, including:
  • the data acquisition module is configured to receive navigation data, ship logs and status information of the unmanned boat;
  • the data processing module is configured to convert the navigation data into the coordinate system and draw a coordinate map, mark the target point and the real-time position of the unmanned boat in the coordinate map, and integrate the ship log and status information;
  • the path planning module is configured to generate the optimal trajectory of the UAV's path planning by adopting the genetic algorithm for data forward driving linear change parameters or the genetic algorithm for data forward driving linear change parameters.
  • the device also includes an autopilot module, which is configured to integrate the optimal trajectory of the unmanned boat's path planning with ship log and status information to obtain the unmanned boat's trajectory and the distance and deflection angle from the target point for control
  • the steering gear performs track offset correction and ship speed adjustment.
  • the receipt data acquisition module receives the heading direction and position data obtained by the navigation data processing subsystem using various sensors such as electronic compass and GPS sensor. Navigation data is stored in real time together with ship log and status information. Then, all the information is processed and passed to the path planning module, where the genetic algorithm is applied to generate the optimal trajectory. According to the planned waypoint, the autopilot uses a closed-loop controller to determine the heading and speed of the USV. In addition, a graphical user interface compiler based on the Spring MVC framework is used to process and record all data in the personal computer.
  • the GPRS wireless network is used as the communication unit between the USV and the personal computer, the effective distance is 5 kilometers, and the transmission speed is 1-100Mbps.
  • the USV containing the device is used for testing, and the overall size and the maximum number of iterations of each algorithm are set to 100 and 1000, respectively.
  • the planning point considers four numbers corresponding to the four working conditions: 15, 25, 35, and 45.
  • the test site is located near the Qingdao Olympic Sailing Center in Fushan Bay. It should be noted that all working conditions consider the same starting point (N36°03'22.38", E120°22'57.06").
  • the best trajectories generated by different algorithms and different planning points are given.
  • the abscissa and ordinate respectively represent the latitude and longitude value of each planning point.
  • the red numbers indicate the order of randomly generated points, corresponding to the order of Appendix I-IV.
  • the starting point is enclosed in a red rectangle, and the arrow indicates the direction of the planned path.
  • Table 4 lists the specific quantitative values of the trajectory.
  • one or more embodiments of the present disclosure propose data forward-driven linear change parameter genetic algorithm (LCPGA) and data reverse In order to compare their path planning performance with traditional algorithms, Monte Carlo simulation was carried out and applied to USV.
  • LCPGA linear change parameter genetic algorithm
  • Monte Carlo simulation shows that the number of planning points has a significant impact on the mean and variance of the optimal path distance of all algorithms, but the impact of population size is limited and can be ignored. Compared with the CGA algorithm, the two improved algorithms can effectively avoid local optima, maintain a higher population diversity in the subsequent iteration process, reduce the average optimal distance, and have better robustness, especially In the case of more planning points.
  • the contribution of one or more embodiments of the present disclosure is to optimize the CGA based on the probability of a time-varying genetic operator to reduce the path distance and enhance the robustness of the algorithm.
  • the new algorithm proposed, especially RLCPGA still has great application prospects.
  • Combining the influence of wind, waves and currents in the actual marine environment on the USV path planning is the main content of the next step.
  • the performance of the obstacle avoidance module can be improved.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Biophysics (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Marketing (AREA)
  • General Health & Medical Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Artificial Intelligence (AREA)
  • Operations Research (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Development Economics (AREA)
  • Molecular Biology (AREA)
  • Game Theory and Decision Science (AREA)
  • Evolutionary Computation (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Data Mining & Analysis (AREA)
  • Software Systems (AREA)
  • Quality & Reliability (AREA)
  • Computational Linguistics (AREA)
  • Biomedical Technology (AREA)
  • Genetics & Genomics (AREA)
  • Physiology (AREA)
  • Automation & Control Theory (AREA)
  • Primary Health Care (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统,该方法包括:接收无人艇的导航数据、船舶日志和状态信息(S1);将导航数据转换坐标系并绘制坐标图,在坐标图中标记目标点和无人艇实时位置(S2);采用数据正向驱动线性变化参数遗传算法或数据反向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹(S3);将无人艇路径规划的最优轨迹整合船舶日志和状态信息,获得无人艇的航行轨迹以及与目标点之间的距离和偏转角度,用于控制舵机进行航迹偏移修正及船速调整。

Description

一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统 技术领域
本公开属于无人艇路径规划的技术领域,涉及一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统。
背景技术
本部分的陈述仅仅是提供了与本公开相关的背景技术信息,不必然构成在先技术。
近年来,无人艇(USV)的发展引起了全世界的关注,无人艇的路径规划问题是一个典型的NP困难问题。智能算法是解决无人艇的路径规划问题的有效途径,其中遗传算法,模拟退火算法,蚁群算法、灰狼优化算法和神经网络算法等启发式算法的研究是现阶段发展的主流。
与其他智能算法相比,遗传算法(GA)允许在结构对象上直接操作,而无需对导数和函数连续性进行限制的特点使得该算法具有更便捷的操纵性和更好的全局搜索能力。此外,GA避免了固化的通过利用率获取最佳搜索空间和搜索方向的问题。由于遗传算法的快速性和高质量,其已成为解决NP难题的最有效方法之一。
然而,发明人在研究过程中发现,传统的遗传算法(CGA)在解决无人艇的路径规划问题中存在遗传算法收敛速度慢、计算效率低、容易出现早熟收敛(局部最优)的问题。
发明内容
针对现有技术中存在的不足,本公开的一个或多个实施例提供了一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统,通过数据正向驱动线性变参数遗传算法或数据反向驱动线性变参数遗传算法,自适应动然调整交叉概率和变异概率,有效快速遍历所有候选方案,在更广阔的领域找到最优路径,同时具有较短的最优路径、较快的收敛速度和较好的鲁棒性。
根据本公开的一个或多个实施例的一个方面,提供一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法。
一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,该方法包括:
接收无人艇的导航数据、船舶日志和状态信息;
将导航数据转换坐标系并绘制坐标图,在坐标图中标记目标点和无人艇实时位置;
采用数据正向驱动线性变化参数遗传算法或数据正向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹。
进一步地,无人艇的所述导航数据为无人艇导航系统采集的经纬度数值,所述将导航数据转换坐标系并绘制坐标图的具体步骤包括:
将无人艇导航系统采集的经纬度数值转换为直角坐标系坐标值;
基于直角坐标系绘制当前海面环境坐标图。
进一步地,在该方法中,所述采用数据正向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹的具体步骤包括:
进行参数初始化,所述参数包括种群规模、最大迭代次数、当前迭代次数、交叉概率和变异概率;
确定染色体长度并对任务序列进行编码生成初始种群;
计算适应度值,并对初始种群进行排序;
计算自适应交叉概率和变异概率,随着迭代次数的增加,交叉概率从0.9线性递减到0.1,变异概率从0.001线性递增到0.1;
根据自适应交叉概率和变异概率对染色体进行自适应交叉、变异操作;
判断是否满足迭代终止条件;
若满足迭代终止条件,从各迭代保留结果中选择最优个体作为正向驱动线性变化参数遗传算法的最优解,否则继续迭代。
进一步地,在该方法中,根据当前迭代次数与交叉概率最终值和初始值差值的乘积除以最大迭代次数与交叉概率的初始值的和计算自适应交叉概率;
根据当前迭代次数与变异概率最终值和初始值差值的乘积除以最大迭代次数与变异概率的初始值的和计算自适应变异概率。
进一步地,在该方法中,所述采用数据反向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹的具体步骤包括:
进行参数初始化,所述参数包括种群规模、最大迭代次数、当前迭代次数、交叉概率和变异概率;
确定染色体长度并对任务序列进行编码生成初始种群;
计算适应度值,并对初始种群进行排序;
计算自适应交叉概率和变异概率,随着迭代次数的增加,交叉概率从0.1线性增加到0.9,而变异概率从0.001线性递增到0.1;
根据自适应交叉概率和变异概率对染色体进行自适应交叉、变异操作;
判断是否满足迭代终止条件;
若满足迭代终止条件,从各迭代保留结果中选择最优个体作为正向驱动线性变化参数遗传算法的最优解,否则继续迭代。
进一步地,在该方法中,根据最大迭代次数和当前迭代次数差值与交叉概率最终值和初始值差值的乘积除以最大迭代次数与交叉概率的初始值的和计算自适应交叉概率;
根据当前迭代次数与变异概率最终值和初始值差值的乘积除以最大迭代次数与变异概率的初始值的和计算自适应变异概率。
进一步地,该方法还包括,将无人艇路径规划的最优轨迹整合船舶日志和状态信息,获得无人艇的航行轨迹以及与目标点之间的距离和偏转角度,用于控制舵机进行航迹偏移修正及船速调整。
根据本公开的一个或多个实施例的一个方面,提供一种计算机可读存储介质。
一种计算机可读存储介质,其中存储有多条指令,所述指令适于由终端设备的处理器加载并执行所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法。
根据本公开的一个或多个实施例的一个方面,提供一种终端设备。
一种终端设备,其包括处理器和计算机可读存储介质,处理器用于实现各指令;计算机可读存储介质用于存储多条指令,所述指令适于由处理器加载并执行所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法。
根据本公开的一个或多个实施例的一个方面,提供一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划装置。
一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划装置,基于所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,包括:
数据获取模块,被配置为接收无人艇的导航数据、船舶日志和状态信息;
数据处理模块,被配置为将导航数据转换坐标系并绘制坐标图,在坐标图中标记目标点和无人艇实时位置,整合船舶日志和状态信息;
路径规划模块,被配置为采用数据正向驱动线性变化参数遗传算法或数据正向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹。
公开的有益效果:
本公开提供的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统,克服传统遗传算法收敛速度慢、计算效率低、容易出现早熟收敛(局部最优)等缺点,与CGA算法相比,两种改进算法都能有效地避免局部最优,在以后的迭代过程中保持更高的种群多样性,减少平均最优距离,并具有更好的鲁棒性,特别是对于规划点较多的情况下。
附图说明
构成本申请的一部分的说明书附图用来提供对本申请的进一步理解,本申请的示意性实施例及其说明用于解释本申请,并不构成对本申请的不当限定。
图1是本公开一个或多个实施例提供的一种基于数据正向驱动线性变参数遗传算法的无人艇路径规划方法流程图;
图2是本公开一个或多个实施例提供的遗传算法流程图;
图3是本公开一个或多个实施例提供的五种待规划点数下的最优路径对图;
图4是本公开一个或多个实施例提供的最优路径距离与待规划点数的平均路径距离图;
图5是本公开一个或多个实施例提供的五最优路径距离与待规划点数的最优路径距离的方差图;
图6是本公开一个或多个实施例提供的最佳路径距离与具有五种种群数量的蒙特卡罗模拟数量的比较图;
图7是本公开一个或多个实施例提供的最佳路径距离与总体大小的平均值比较图;
图8是本公开一个或多个实施例提供的最佳路径距离与总体大小的平均值和方差值比较图;
图9是本公开一个或多个实施例提供的50个规划点下三种算法的收敛曲线图;
图10是本公开一个或多个实施例提供的50个规划点下三种算法的最优路径图;
图11是本公开一个或多个实施例提供的USV测试时四个规划点的最佳轨迹图。
具体实施方式:
下面将结合本公开的一个或多个实施例中的附图,对本公开的一个或多个实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本公开一部分实施例,而不是全部的实施例。基于本公开的一个或多个实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本公开保护的范围。
应该指出,以下详细说明都是例示性的,旨在对本申请提供进一步的说明。除非另有指明,使用的所有技术和科学术语具有与本申请所属技术领域的普通技术人员通常理解的相同含义。
需要注意的是,这里所使用的术语仅是为了描述具体实施方式,而非意图限制根据本申请的示例性实施方式。如在这里所使用的,除非上下文另外明确指出,否则单数形式也意图包括复数形式,此外,还应当理解的是,当在本说明书中使用术语“包含”和/或“包括”时,其指明存在特征、步骤、操作、器件、组件和/或它们的组合。
需要注意的是,附图中的流程图和框图示出了根据本公开的各种实施例的方法和系统的可能实现的体系架构、功能和操作。应当注意,流程图或框图中的每个方框可以代表一个模块、程序段、或代码的一部分,所述模块、程序段、或代码的一部分可以包括一个或多个用于实现各个实施例中所规定的逻辑功能的可执行指令。也应当注意,在有些作为备选的实现中,方框中所标注的功能也可以按照不同于附图中所标注的顺序发生。例如,两个接连地表示的方框实际上可以基本并行地执行,或者它们有时也可以按照相反的顺序执行,这取决于所涉及的功能。同样应当注意的是,流程图和/或框图中的每个方框、以及流程图和/或框图中的方框的组合,可以使用执行规定的功能或操作的专用的基于硬件的系统来实现,或者可以使用专用硬件与计算机指令的组合来实现。
在不冲突的情况下,本公开中的实施例及实施例中的特征可以相互组合,下面结合附图与实施例对本公开作进一步说明。
传统遗传算法(CGA):传统遗传算法(CGA)选择实数编码的方法,使用带有访问城市序列号的字符串来表示每条染色体。遗传参数,例如种群大小、交叉和突变概率,通常据经验定义。在确定优化问题后,随机生成具有一定规模的候选解的初始总体。适应性函数为1/len(len代表每条染色体的相对路径长度),用于评估每个个体的适应性,更适合的个体将在繁殖中存活下来。然后,该算法通过交叉、变异和选择的迭代操作来提高种群数量,如果满足某个标准或达到最大迭代次数,则进化过程将终止。
在CGA中,交叉主要用于连接两条父染色体,这些染色体被确定的断裂点分开,并产生两个具有一定交叉概率(PC)的后代。突变主要用于交换染色体上随机选择的两个突变点上的基因位置,且突变的发生具有一定的突变概率(PM)。应该指出的是,交叉使染色体相似,有助于种群的收敛;而突变增大遗传多样性,使算法在局部最优的情况下能进一步扩展种群数。
实施例一
如图1所示,一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,该方法包括:
一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,该方法包括:
步骤S1:接收无人艇的导航数据、船舶日志和状态信息;
步骤S2:将导航数据转换坐标系并绘制坐标图,在坐标图中标记目标点和无人艇实时位置;
步骤S3:采用数据正向驱动线性变化参数遗传算法或数据正向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹。
步骤S4:将无人艇路径规划的最优轨迹整合船舶日志和状态信息,获得无人艇的航行轨迹以及与目标点之间的距离和偏转角度,用于控制舵机进行航迹偏移修正及船速调整。
在本实施例的步骤S1中,无人艇的所述导航数据为无人艇导航系统采集的经纬度数值。
在本实施例的步骤S2中,所述将导航数据转换坐标系并绘制坐标图的具体步骤包括:
将无人艇导航系统采集的经纬度数值转换为直角坐标系坐标值;
基于直角坐标系绘制当前海面环境坐标图。
在本实施例中,为了避免CGA中P C和P M控制参数确定的主观性,对A.Ratnaweera等人提出的“时变加速度系数”概念进行了改进。在最初的迭代中使用了一个较大的个体认知系数和一个较小的社会认知系数,以允许粒子尽可能的搜索更大空间,避免算法收敛到局部最优。随着时间的推移,个体认知系数减少,而社会认知系数增加,以保证算法能成功收敛在可控区间内,避免算法遭遇无法收敛的情况。
在本实施例的步骤S3中,遗传算法的遗传算子参数随迭代次数线性变化。根据数据驱动的方向,改进分为两类,即正向型(LCPGA)和反向型(RLCPGA)。这两种类型都可以帮助个体遍历所有候选解,并在广泛的领域中找到最优解。
在LCPGA中,P C和P M的初始值分别在0.1-0.9和0.001-0.1范围内变化。随着迭代次数的增加,P C从0.9线性递减到0.1,而P M从0.001线性递增到0.1。方程式(1)和(2)分别表示更新P C和P M值的数学相关性:
P C(k)=P cs+k*(P ce-P cs)/MAXDEN     (1)
P m(k)=P ms+k*(P me-P ms)/MAXDEN    (2)
其中下标e和s分别表示P C和P M的最终值和初始值,k表示当前的迭代数,MAXGEN表示最大迭代次数。图2显示了改进的算法的在流程中的具体作用位置。
所述采用数据正向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹的具体步骤包括:
步骤S3-1:进行参数初始化,所述参数包括种群规模、最大迭代次数、当前迭代次数、交叉概率和变异概率;设置种群规模N、最大迭代次数MAXGEN、当前迭代次数k、初始交叉概率P c、初始变异概率P m
步骤S3-2:确定染色体长度并对任务序列进行编码生成初始种群;采用实数形式编码任务序列,随机产生U中的N个染色体S 1,S 2,...,S N组成初始种群S={S 1,S 2,...,S N}置代数计数器g=1;
步骤S3-3:计算适应度值,并对初始种群进行排序;适应度函数定义为1/len,其中len代表每条染色体的相对路径长度。根据计算所得适应度值对初始种群进行排序。
步骤S3-4:计算自适应交叉概率和变异概率,随着迭代次数的增加,交叉概率从0.9线性递减到0.1,变异概率从0.001线性递增到0.1;
设计了一种自适应交叉、变异算子,因为在算法初期需要产生足够多的新染色体以保证最优个体的存在概率,此时选择较大的P C可以帮助产生新的个体;当算法进行到后期时,种群中个体适应度趋于一致时,种群容易陷入局部最优,此时种群多样性较差,选择较大的P M可以增加新个体产生的概率,有利于种群跳出局部最优。因此,根据当前迭代次数与交叉概率最终值和初始值差值的乘积除以最大迭代次数与交叉概率的初始值的和计算自适应交叉概率;根据当前迭代次数与变异概率最终值和初始值差值的乘积除以最大迭代次数与变异概率的初始值的和计算自适应变异概率。采用下式具体计算自适应交叉概率和变异概率:
P c(k)=P cs+k*(P ce-P cs)/MAXDEN    (1)
P m(k)=P ms+k*(P me-P ms)/MAXDEN    (2)
式中,下标e和s分别表示P C和P M的最终值和初始值,k表示当前的迭代数,MAXGEN表示最大迭代次数。
步骤S3-5:根据自适应交叉概率和变异概率对染色体进行自适应交叉、变异操作;计算所得到的交叉概率和变异概率后,对染色体进行自适应交叉、变异操作。为保证每条染色体中的基因码出现且仅出现一次,交叉算子采用循环交叉算子,变异算子采用基因对换变异算子。计算新子代的适应度值,根据该值重新排序,获得新的种群进入终止条件判断。
步骤S3-6:判断是否满足迭代终止条件;迭代终止条件设定为满足某一工况需求或迭代次数达到最大。若不满足终止条件,则迭代次数加一,转步骤S3-3;若满足则转步骤S3-7。
步骤S3-7:从各迭代保留结果中选择最优个体作为正向驱动线性变化参数遗传算法的最优解并输出,整个算法结束。
研究发现,一旦一个优秀个体出现在LCPGA中,随着迭代的进行,群体中的其他个体有接近优秀个体的倾向。同时,当迭代接近尾声时,个体之间的差异不断减小,P C处于一个较低的水平,一定程度上降低了算法的种群多样性。如果此时的最优个体是局部最优,则该算法最终可能达到局部最优,这种现象对于多模态函数的路径规划问题尤为显著。
通过对交叉概率的调整,较高的P C可能产生更多的“子”染色体,从而增强了局部搜索的能力,提高了算法向全局最优收敛的能力。另一方面,较低的P C可以直接将“父”染色体 转换为“子”染色体,而无需交叉操作,这有助于保持“父”代的多样性,并确保在初始迭代期间具有理想的全局搜索能力。
因此,基于时变动态控制P C的概念提出了数据反向驱动型遗传算法(RLCPGA),如图2所示,P C和P M的初始值分别设置在0.1~0.9和0.001~0.1范围内。随着迭代的进行,P C从0.1线性增加到0.9,P M从0.001线性增加到0.1。理论上,一个较小的P C和一个较大的P M在开始迭代时将有助于保持初始群体的多样性,并帮助个体遍历所有候选解决方案。同时,在迭代后期,一个较大的P C和一个较小的P M将有助于个体以更快的收敛速度围绕全局最优点进行集群。
P C和P M在每次迭代中的更新公式如下所示:
P c(k)=P cs+(MAXGEN-k)*(P ce-P cs)/MAXGEN        (3)
P m(k)=P ms+k*(P me-P ms)/MAXGEN        (4)
式(3)表明,根据迭代序列号,P C值可以自适应优化。为了保护随机生成群体的多样性,提高全局搜索的能力,P C在初始迭代中是很小的,当所有个体都趋向于与不断增加的迭代保持一致时,P C会增加,局部搜索的能力随之增强。
在本实施例的步骤S3中,所述采用数据反向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹的具体步骤包括:
C1、参数初始化。设置种群规模N、最大迭代次数MAXGEN、当前迭代次数k、初始交叉概率P c、初始变异概率P m
C2、确定染色体长度并对任务序列进行编码生成初始种群;采用实数形式编码任务序列,随机产生U中的N个染色体S 1,S 2,...,S N组成初始种群S={S 1,S 2,...,S N}置代数计数器g=1;
C3、适应度值的计算。适应度函数定义为1/len,其中len代表每条染色体的相对路径长度。根据计算所得适应度值对初始种群进行排序。
C4、对染色体进行自适应交叉、变异操作。通过对交叉概率的调整,较高的P C可能产生更多的“子”染色体,从而增强局部搜索的能力,提高了算法向全局最优收敛的能力。另一方面,较低的P C可以直接将“父”染色体转换为“子”染色体,而无需交叉操作,这有助于保持“父”代的多样性,并确保在初始迭代期间具有理想的全局搜索能力。因此,基于时变动态控制P C的概念提出了数据反向驱动型遗传算法(RLCPGA),P C和P M的初始值分别设置在0.1~0.9和0.001~0.1范围内。随着迭代的进行,根据迭代序列号,P C值可以自适应优化。为了保护随机生成群体的多样性,提高全局搜索的能力,P C在初始迭代中是很小的,当所有个体都趋向于与不断增加的迭代保持一致时,P C会增加,局部搜索的能力随之增强。
根据最大迭代次数和当前迭代次数差值与交叉概率最终值和初始值差值的乘积除以最大迭代次数与交叉概率的初始值的和计算自适应交叉概率;
根据当前迭代次数与变异概率最终值和初始值差值的乘积除以最大迭代次数与变异概率的初始值的和计算自适应变异概率。
P C和P M在每次迭代中的更新公式如下所示:
P c(k)=P cs+(MAXGEN-k)*(P ce-P cs)/MAXGEN        (3)
P m(k)=P ms+k*(P me-P ms)/MAXGEN        (4)
式中下标e和s分别表示P C和P M的最终值和初始值,k表示当前的迭代数,MAXGEN表示最大迭代次数。
C5、计算所得到的交叉概率和变异概率后,对染色体进行自适应交叉、变异操作。为保证每条染色体中的基因码出现且仅出现一次,交叉算子采用循环交叉算子,变异算子采用基因对换变异算子。计算新子代的适应度值,根据该值重新排序,获得新的种群进入终止条件判断。
C6、迭代终止条件判断。迭代终止条件设定为满足某一工况需求或迭代次数达到最大。若不满足终止条件,则迭代次数加一,转C3;若满足则转C7。
C7、从各迭代保留结果中选择最优个体作为反向驱动线性变化参数遗传算法的最优解并输出,整个算法结束。
表1.五种待规划点数下最优路径的均值和方差
Figure PCTCN2019112603-appb-000001
如表1所示,从规划点的数量角度比较了CGA、LCPGA和RLCPGA对路径规划的影响。规划点数(P)考虑了5个不同种群大小的方案,具体种群数方案如:10、20、30、40和50。此外,每个算法的总体大小S和最大迭代次数分别设置为100和1000。为了获得路径规划的结果,对每个算法和工作条件进行了100次蒙特卡罗模拟实验。
图3显示了针对不同算法的蒙特卡罗模拟次数(N)和最佳路径距离(D)的分布。其中左列、中列和右列数字的蓝色符号分别代表CGA、LCPGA和RLCPGA计算所得的数据。此外,100次模拟的平均值在每个图中用红线表示。结果表明,随着P值的增加,符号在平均线上的分布范围越来越大,直线的纵坐标值也越来越大。当p设置为10时,没有明显的差异,而随着P的增加,与CGA相比,LCPGA和RLCPGA的结果更接近平均线。此外,在相同的工作条件下,两种改进算法的平均线明显低于CGA算法的平均线。
如图4和图5所示,给出了不同算法所得最优路径距离的平均值
Figure PCTCN2019112603-appb-000002
和方差(V),从定量角度比较不同算法的路径规划能力。如图4所示,当P为10时,三种算法的平均路径值没有明显差异,说明当规划点数较小时,计算难度较小,三种算法均能稳定计算出当前的最优路径值。当P为20时,CGA计算所得平均值大于另外两种算法,且随着P的增加,CGA与另外两种算法的间隙有增加的趋势。当P为50时,出现最大相对间隙为30%。结果表明,两种优化算法在规划最优路径的距离时,无论规划的点数是多少,其性能基本相同,且均优于传统算法。
计算方差以显示数据集与平均值之间的分布距离,并反映算法的鲁棒性。如图5所示,当P设置为10时,不同算法的方差没有显著差异。此外,在所有工作条件下,LCPGA和RLCPGA的方差曲线存在的差异可以忽略。比较表明,两种改进算法的方差曲线均具有与P相似的增长趋势,其值明显低于CGA。结果表明,当P值为50时,相对间隙最大,且最大相对间隙达到90%。此外,与其他算法相比,RLCPGA的方差曲线保持在最低水平。这意味着RLCPGA是目前研究中最稳定、最可靠的规划点数算法。表1给出了每个算法和每个工作条件的
Figure PCTCN2019112603-appb-000003
和V的详细值。
表2.五种种群数量下最优路径的均值和方差
Figure PCTCN2019112603-appb-000004
如表2所示,将考虑5种种群数量(20、30、60、80和100)下两种改进算法的路径规划性能。每个算法的规划点数和最大迭代次数分别设置为50和1000,对每个算法进行100次蒙特卡罗模拟。
如图6所示,显示了针对不同种群数量(S)在不同模拟数(N)下的最佳路径距离(D)100次蒙特卡罗实验,模拟数据在平均线附近分散。结果表明,对于五种不同的种群数量,改进算法中的模拟数据与CGA中的数据相比更接近于平均线,且LCPGA和RLCPGA算法的平均线明显低于CGA。然而,两种改进算法的数据分散程度和最优路径平均值之间的差异较微小,无法从图5中直接看出,因此,必须根据平均值D和方差V呈现定量结果,如图7和图8所示。
如图7所示,当S=20时,CGA曲线与其他两种改进算法之间存在明显的间隙,其中最大相对间隙为42%。据观察,随着种群数量的增加,差距减小。与图3相似,LCPGA和RLCPGA的平均值几乎没有差异。此外,还发现CGA的平均路径距离明显受种群大小的影响,随着S的增加,其路径距离减小。然而,无论总体大小如何,两种改进算法在规划最优路径的距离方面都具有相似且理想的性能。
如图8所示,各种群规模的CGA方差均保持在0.5m 2以上的最高水平,并随S的增加而下降,同时改进算法的方差曲线具有相似的振荡趋势,但方差值始终小于0.1。研究发现,总体大小对改进算法的影响是不规则的。除S=10外,RLCPGA的方差总是小于LCPGA的方差。例如,当S=100时,改进算法的方差相对差为28%。对于大多数种群数量来说(S=40~100),RLCPGA比其他算法表现出更稳定和可靠的性能。表2给出了每种算法和总体大小的D和V的详细差异。
表3.各算法的计算效率对比
Figure PCTCN2019112603-appb-000005
如表3所示,利用50个随机选择的规划点,比较了三种算法的计算效率。选择了两个评价算法计算效率的主要指标:时间消耗和收敛速度。前者是指完成最大迭代次数所需的时间成本,后者是指解决方案达到收敛水平的临界迭代次数(N cri)。最大迭代次数和种群大小分别设置为1000和100。图9显示了每种算法的最优路径距离与迭代的收敛历史。同时,各算法解的详细信息见表3。
如图9所示,随着迭代次数的增加,每个算法的路径距离(D)逐渐优化为更短,然后在临界数(Ncri)处收敛到稳定水平,最终达到全局最优。相比之下,在整个计算过程中,RLCPGA 的曲线处于最低水平,具有最快的收敛速度和最小的临界数。RLCPGA收敛于N cri=318,比CGA和LCPGA分别快52.5%和6.2%。然而,在完成相同的迭代次数时,RLCPGA和LCPGA的时间消耗分别比CGA高9.8%和9.0%。因此,需要注意的是,改进后的算法大大降低了计算时间成本,实现了最优路径距离的减小,提高了算法的鲁棒性。此外,图10给出了每个算法设计的50个规划点的轨迹。横坐标和纵坐标分别代表每个规划点的经纬度值,起点围在红色矩形内,箭头代表规划路径的走向。可以发现,在图10中,只有CGA生成的轨迹具有明显的路径交叉现象,这就是为什么在相同条件下生成的路径距离比其他算法长的原因。但同时,改进算法的优点在有效避免路径交叉、简化路径形状等方面表现得更为明显。
实施例二
根据本公开的一个或多个实施例的一个方面,提供一种计算机可读存储介质。
一种计算机可读存储介质,其中存储有多条指令,所述指令适于由终端设备的处理器加载并执行所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法。
实施例三
根据本公开的一个或多个实施例的一个方面,提供一种终端设备。
一种终端设备,其包括处理器和计算机可读存储介质,处理器用于实现各指令;计算机可读存储介质用于存储多条指令,所述指令适于由处理器加载并执行所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法。
这些计算机可执行指令在设备中运行时使得该设备执行根据本公开中的各个实施例所描述的方法或过程。
在本实施例中,计算机程序产品可以包括计算机可读存储介质,其上载有用于执行本公开的各个方面的计算机可读程序指令。计算机可读存储介质可以是可以保持和存储由指令执行设备使用的指令的有形设备。计算机可读存储介质例如可以是――但不限于――电存储设备、磁存储设备、光存储设备、电磁存储设备、半导体存储设备或者上述的任意合适的组合。计算机可读存储介质的更具体的例子(非穷举的列表)包括:便携式计算机盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、静态随机存取存储器(SRAM)、便携式压缩盘只读存储器(CD-ROM)、数字多功能盘(DVD)、记忆棒、软盘、机械编码设备、例如其上存储有指令的打孔卡或凹槽内凸起结构、以及上述的任意合适的组合。这里所使用的计算机可读存储介质不被解释为瞬时信号本身,诸如无线电波或者其他自由传播的电磁 波、通过波导或其他传输媒介传播的电磁波(例如,通过光纤电缆的光脉冲)、或者通过电线传输的电信号。
本公开所描述的计算机可读程序指令可以从计算机可读存储介质下载到各个计算/处理设备,或者通过网络、例如因特网、局域网、广域网和/或无线网下载到外部计算机或外部存储设备。网络可以包括铜传输电缆、光纤传输、无线传输、路由器、防火墙、交换机、网关计算机和/或边缘服务器。每个计算/处理设备中的网络适配卡或者网络接口从网络接收计算机可读程序指令,并转发该计算机可读程序指令,以供存储在各个计算/处理设备中的计算机可读存储介质中。
用于执行本公开内容操作的计算机程序指令可以是汇编指令、指令集架构(ISA)指令、机器指令、机器相关指令、微代码、固件指令、状态设置数据、或者以一种或多种编程语言的任意组合编写的源代码或目标代码,所述编程语言包括面向对象的编程语言—诸如C++等,以及常规的过程式编程语言—诸如“C”语言或类似的编程语言。计算机可读程序指令可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中,远程计算机可以通过任意种类的网络—包括局域网(LAN)或广域网(WAN)—连接到用户计算机,或者,可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。在一些实施例中,通过利用计算机可读程序指令的状态信息来个性化定制电子电路,例如可编程逻辑电路、现场可编程门阵列(FPGA)或可编程逻辑阵列(PLA),该电子电路可以执行计算机可读程序指令,从而实现本公开内容的各个方面。
实施例四
根据本公开的一个或多个实施例的一个方面,提供一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划装置。
一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划装置,基于所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,包括:
数据获取模块,被配置为接收无人艇的导航数据、船舶日志和状态信息;
数据处理模块,被配置为将导航数据转换坐标系并绘制坐标图,在坐标图中标记目标点和无人艇实时位置,整合船舶日志和状态信息;
路径规划模块,被配置为采用数据正向驱动线性变化参数遗传算法或数据正向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹。
该装置还包括自动驾驶模块,被配置为将无人艇路径规划的最优轨迹整合船舶日志和状态信息,获得无人艇的航行轨迹以及与目标点之间的距离和偏转角度,用于控制舵机进行航迹偏移修正及船速调整。
收据数据获取模块接收导航数据处理子系统采用电子罗盘和GPS传感器等多种传感器来获取的船首方向和位置数据。导航数据与船舶日志和状态信息一起实时存储。然后,对所有信息进行处理并传递到路径规划模块,在该路径规划模块中应用遗传算法生成最优轨迹。根据计划的航路点,自动驾驶仪采用闭环控制器来确定USV的航向和速度。此外,还采用了基于Spring MVC框架的图形用户界面编译程序,对个人计算机中的所有数据进行处理和记录。以GPRS无线网络作为USV与个人计算机之间的通信单元,有效距离为5公里,传输速度为1-100Mbps。
在本实施例中,通过包含该装置的USV进行测试,每个算法的总体大小和最大迭代次数分别设置为100和1000。此外,在实际环境中,规划点考虑四个对应四种工况的数字,分别为:15、25、35和45。试验场地位于浮山湾青岛奥林匹克帆船中心附近。应指出,所有工作条件均考虑相同的起点(N36°03′22.38″,E120°22′57.06″)。
表4.四种规划点下不同算法的最优路径值
Figure PCTCN2019112603-appb-000006
如图11所示,给出了由不同算法和不同规划点数生成的最佳轨迹。横坐标和纵坐标分别代表每个规划点的经纬度值。红色数字表示随机生成点的顺序,与附录Ⅰ-Ⅳ的顺序相对应。起点封闭在红色矩形中,箭头表示规划路径的方向。此外,为了更好地分析算法的路径规划性能,表4列出了轨迹的具体量化值。
如图11所示,对于15个规划点,各个算法的路径形状和数值没有显著差异。随着规划点数的增加,轨迹变得更加复杂,路径形状和数值的差异也增大。当P=25时,在序号为3、6、11和15的点附近观察到明显差异。当P值超过25时,CGA产生的轨迹有交叉现象,该现象在某种程度上导致了更长的路径值,而改进算法在有效避免路径交叉和简化路径形状方面的优势 随着P的增加更加明显。进一步的研究表明,当P值大于25时,RLCPGA在减小路径距离方面的性能优于LCPGA。此外,改进算法之间的最大差异出现在P=35时,达到路径距离的3%。
应当注意,尽管在上文的详细描述中提及了设备的若干模块或子模块,但是这种划分仅仅是示例性而非强制性的。实际上,根据本公开的实施例,上文描述的两个或更多模块的特征和功能可以在一个模块中具体化。反之,上文描述的一个模块的特征和功能可以进一步划分为由多个模块来具体化。
为了克服传统遗传算法收敛速度慢、计算效率低、容易出现早熟收敛(局部最优)等缺点,本公开一个或多个实施例提出了数据正向驱动线性变化参数遗传算法(LCPGA)和数据反向驱动线性变化参数遗传算法(RLCPGA)两种自适应调整参数的改进遗传算法,为了比较它们的路径规划性能与传统算法,进行了蒙特卡罗模拟,并将其应用于USV。
蒙特卡罗模拟表明,规划点的个数对所有算法的最优路径距离的均值和方差有明显的影响,但是种群数量的影响是有限的,可以忽略不计。与CGA算法相比,两种改进算法都能有效地避免局部最优,在以后的迭代过程中保持更高的种群多样性,减少平均最优距离,并具有更好的鲁棒性,特别是对于规划点较多的情况下。
最后将所提出的算法应用于实际场地的USV路径规划问题。结果表明,当P值超过25时,CGA会产生一个有交叉现象的轨迹,这会增加路径形状和距离的复杂性,改进后的算法可以有效地避免这一现象,减少路径距离。进一步比较表明,RLCPGA在USV规划路径方面优于其它算法。
本公开一个或多个实施例的贡献在于基于时变遗传算子概率对CGA进行优化,以减少路径距离,增强算法的鲁棒性。虽然本文的研究工作提供了初步的研究,但所提出的新算法,特别是RLCPGA仍具有很大的应用前景。结合实际海洋环境中的风、浪、流对USV路径规划的影响问题是下一步研究的主要内容。同时,通过提高导航数据的质量和可靠性,可以提高避障模块的性能。
以上所述仅为本申请的优选实施例而已,并不用于限制本申请,对于本领域的技术人员来说,本申请可以有各种更改和变化。凡在本申请的精神和原则之内,所作的任何修改、等同替换、改进等,均应包含在本申请的保护范围之内。因此,本公开将不会被限制于本公开所示的这些实施例,而是要符合与本公开所公开的原理和新颖特点相一致的最宽的范围。

Claims (10)

  1. 一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,其特征在于,该方法包括:
    接收无人艇的导航数据、船舶日志和状态信息;
    将导航数据转换坐标系并绘制坐标图,在坐标图中标记目标点和无人艇实时位置;
    采用数据正向驱动线性变化参数遗传算法或数据正向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹。
  2. 如权利要求1所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,其特征在于,无人艇的所述导航数据为无人艇导航系统采集的经纬度数值,所述将导航数据转换坐标系并绘制坐标图的具体步骤包括:
    将无人艇导航系统采集的经纬度数值转换为直角坐标系坐标值;
    基于直角坐标系绘制当前海面环境坐标图。
  3. 如权利要求1所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,其特征在于,在该方法中,所述采用数据正向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹的具体步骤包括:
    进行参数初始化,所述参数包括种群规模、最大迭代次数、当前迭代次数、交叉概率和变异概率;
    确定染色体长度并对任务序列进行编码生成初始种群;
    计算适应度值,并对初始种群进行排序;
    计算自适应交叉概率和变异概率,随着迭代次数的增加,交叉概率从0.9线性递减到0.1,变异概率从0.001线性递增到0.1;
    根据自适应交叉概率和变异概率对染色体进行自适应交叉、变异操作;
    判断是否满足迭代终止条件;
    若满足迭代终止条件,从各迭代保留结果中选择最优个体作为正向驱动线性变化参数遗传算法的最优解,否则继续迭代。
  4. 如权利要求3所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,其特征在于,在该方法中,根据当前迭代次数与交叉概率最终值和初始值差值的乘积除以最大迭代次数与交叉概率的初始值的和计算自适应交叉概率;
    根据当前迭代次数与变异概率最终值和初始值差值的乘积除以最大迭代次数与变异概率的初始值的和计算自适应变异概率。
  5. 如权利要求1所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,其特征在于,在该方法中,所述采用数据反向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹的具体步骤包括:
    进行参数初始化,所述参数包括种群规模、最大迭代次数、当前迭代次数、交叉概率和变异概率;
    确定染色体长度并对任务序列进行编码生成初始种群;
    计算适应度值,并对初始种群进行排序;
    计算自适应交叉概率和变异概率,随着迭代次数的增加,交叉概率从0.1线性增加到0.9,而变异概率从0.001线性递增到0.1;
    根据自适应交叉概率和变异概率对染色体进行自适应交叉、变异操作;
    判断是否满足迭代终止条件;
    若满足迭代终止条件,从各迭代保留结果中选择最优个体作为正向驱动线性变化参数遗传算法的最优解,否则继续迭代。
  6. 如权利要求5所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,其特征在于,在该方法中,根据最大迭代次数和当前迭代次数差值与交叉概率最终值和初始值差值的乘积除以最大迭代次数与交叉概率的初始值的和计算自适应交叉概率;
    根据当前迭代次数与变异概率最终值和初始值差值的乘积除以最大迭代次数与变异概率的初始值的和计算自适应变异概率。
  7. 如权利要求1所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,其特征在于,该方法还包括,将无人艇路径规划的最优轨迹整合船舶日志和状态信息,获得无人艇的航行轨迹以及与目标点之间的距离和偏转角度,用于控制舵机进行航迹偏移修正及船速调整。
  8. 一种计算机可读存储介质,其中存储有多条指令,其特征在于,所述指令适于由终端设备的处理器加载并执行如权利要求1-7任一项所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法。
  9. 一种终端设备,其包括处理器和计算机可读存储介质,处理器用于实现各指令;计算机可读存储介质用于存储多条指令,其特征在于,所述指令适于由处理器加载并执行如权利要求1-7任一项所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法。
  10. 一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划装置,其特征在于,基于如权利要求1-7任一项所述的一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法,包括:
    数据获取模块,被配置为接收无人艇的导航数据、船舶日志和状态信息;
    数据处理模块,被配置为将导航数据转换坐标系并绘制坐标图,在坐标图中标记目标点和无人艇实时位置,整合船舶日志和状态信息;
    路径规划模块,被配置为采用数据正向驱动线性变化参数遗传算法或数据正向驱动线性变化参数遗传算法生成无人艇路径规划的最优轨迹。
PCT/CN2019/112603 2019-08-28 2019-10-22 一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统 WO2021035911A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201910802194.XA CN110516877A (zh) 2019-08-28 2019-08-28 一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统
CN201910802194.X 2019-08-28

Publications (1)

Publication Number Publication Date
WO2021035911A1 true WO2021035911A1 (zh) 2021-03-04

Family

ID=68627729

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/112603 WO2021035911A1 (zh) 2019-08-28 2019-10-22 一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统

Country Status (3)

Country Link
CN (1) CN110516877A (zh)
LU (1) LU101989B1 (zh)
WO (1) WO2021035911A1 (zh)

Cited By (31)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113191086A (zh) * 2021-05-08 2021-07-30 国网山东省电力公司营销服务中心(计量中心) 一种基于遗传算法的电采暖热负荷需求优化方法及系统
CN113517034A (zh) * 2021-05-21 2021-10-19 华东理工大学 分子筛工艺条件的生成方法和系统
CN113625767A (zh) * 2021-09-02 2021-11-09 大连海事大学 一种基于优选信息素灰狼算法的固定翼无人机集群协同路径规划方法
CN113630737A (zh) * 2021-08-04 2021-11-09 西安电子科技大学 一种无线可充电传感器网络中移动充电器的部署方法
CN113673782A (zh) * 2021-09-06 2021-11-19 中导光电设备股份有限公司 多显微镜扫描拍照路径优化方法和装置
CN113741477A (zh) * 2021-09-14 2021-12-03 中国船舶重工集团公司第七0七研究所九江分部 一种欠驱动船舶靠泊路径规划方法
CN113759915A (zh) * 2021-09-08 2021-12-07 广州杰赛科技股份有限公司 一种agv小车路径规划方法、装置、设备及存储介质
CN113778128A (zh) * 2021-09-18 2021-12-10 中国人民解放军国防科技大学 基于遗传算法的无人机侦察任务规划方法及装置
CN113780745A (zh) * 2021-08-16 2021-12-10 华中科技大学 一种上门服务需求驱动的it人员调度方法和系统
CN113867358A (zh) * 2021-10-13 2021-12-31 中国人民解放军陆军工程大学 多无人车协同遍历任务的智能路径规划方法
CN113901085A (zh) * 2021-09-30 2022-01-07 中远海运科技股份有限公司 一种船舶轨迹动态绘制方法及系统
CN113985899A (zh) * 2021-11-25 2022-01-28 江苏科技大学 基于区间多目标优化的水下机器人全局路径规划方法
CN114035587A (zh) * 2021-11-24 2022-02-11 陕西欧卡电子智能科技有限公司 一种无人船集群多船协同的路径规划方法、装置及无人船
CN114158105A (zh) * 2021-11-08 2022-03-08 石河子大学 应用于综合能源系统的无线传感器网络的路由方法及装置
CN114282435A (zh) * 2021-12-16 2022-04-05 青岛科技大学 一种基于改进遗传算法的无人船航线规划方法
CN114545621A (zh) * 2022-03-15 2022-05-27 中国科学院光电技术研究所 一种基于遗传算法的变形反射镜结构优化方法
CN114611243A (zh) * 2022-03-14 2022-06-10 华中科技大学 盾构通用型管片点位选取模型构建方法及点位选取方法
CN114818465A (zh) * 2022-03-16 2022-07-29 西北工业大学 一种基于nurbs曲面的低散射优化方法
CN115079704A (zh) * 2022-08-01 2022-09-20 中国电信股份有限公司 路径规划方法、装置、存储介质及电子设备
CN115712507A (zh) * 2022-11-29 2023-02-24 上海船舶运输科学研究所有限公司 一种计算船舶网关任务优先级的方法
CN115857516A (zh) * 2023-03-02 2023-03-28 之江实验室 结合牛耕式运动与遗传算法的全覆盖路径规划方法和装置
CN116402317A (zh) * 2023-06-05 2023-07-07 山东梧桐城市规划技术服务有限公司 一种城市环境大数据综合管理系统
CN116449685A (zh) * 2023-04-23 2023-07-18 上海交通大学 一种快速跟踪精准控制的舰船燃机控制方法及装置
CN116678409A (zh) * 2023-04-24 2023-09-01 大连海事大学 一种基于改进灰狼算法的北极冰区船舶路径规划方法
CN116707624A (zh) * 2023-08-09 2023-09-05 北京天航华创科技股份有限公司 一种多艇组网通导遥系统及方法
CN117034480A (zh) * 2023-07-07 2023-11-10 北京航空航天大学 基于参数化降维的压气机多型面一体化优化方法及平台
CN117196019A (zh) * 2023-07-27 2023-12-08 南京中禹智慧水利研究院有限公司 基于改进自适应遗传算法的新安江模型参数率定方法
CN117252114A (zh) * 2023-11-17 2023-12-19 湖南华菱线缆股份有限公司 一种基于遗传算法的电缆耐扭转实验方法
CN117455918A (zh) * 2023-12-25 2024-01-26 深圳市辉熙智能科技有限公司 基于图像分析的防堵送料方法及系统
CN117689677A (zh) * 2024-02-01 2024-03-12 山东大学日照研究院 一种砂轮磨损状态识别方法、系统、设备及介质
CN117806328A (zh) * 2023-12-28 2024-04-02 华中科技大学 一种基于基准标记的无人艇靠泊视觉引导控制方法及系统

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
RU2735694C1 (ru) * 2020-05-10 2020-11-05 Федеральное государственное бюджетное образовательное учреждение высшего образования «Государственный университет морского и речного флота имени адмирала С.О. Макарова» Способ натурных испытаний безэкипажных судов
CN112035921B (zh) * 2020-08-25 2021-06-29 中船文化科技(北京)有限公司 一种方舱动线规划方法、装置、电子设备以及存储介质
CN113095578B (zh) * 2021-04-16 2022-09-23 广东电网有限责任公司电力调度控制中心 变电站最优通信路径的设计方法、装置、终端及介质
RU2760823C1 (ru) * 2021-05-04 2021-11-30 Федеральное государственное бюджетное образовательное учреждение высшего образования "Государственный морской университет имени адмирала Ф.Ф. Ушакова" Опытовый морской модульный комплекс
CN113959448B (zh) * 2021-10-26 2023-08-29 江苏海洋大学 一种基于改进樽海鞘群算法的水下地形辅助导航方法
CN114897246A (zh) * 2022-05-16 2022-08-12 北京航空航天大学 基于遗传算法和贪心策略的飞艇多任务点航线规划方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6052678A (en) * 1993-12-16 2000-04-18 Fujitsu Limited Problem solving operation apparatus using a state transition
CN106845716A (zh) * 2017-01-25 2017-06-13 东南大学 一种基于导航误差约束的水面无人艇局部分层路径规划方法
CN108388250A (zh) * 2018-03-30 2018-08-10 哈尔滨工程大学 一种基于自适应布谷鸟搜索算法的水面无人艇路径规划方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6052678A (en) * 1993-12-16 2000-04-18 Fujitsu Limited Problem solving operation apparatus using a state transition
CN106845716A (zh) * 2017-01-25 2017-06-13 东南大学 一种基于导航误差约束的水面无人艇局部分层路径规划方法
CN108388250A (zh) * 2018-03-30 2018-08-10 哈尔滨工程大学 一种基于自适应布谷鸟搜索算法的水面无人艇路径规划方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
JUNFENG XIN, YONGBO ZHANG, JIAGENG BO, BOWEN ZHAO, SHIYUAN FAN: "Study on path planning of unmanned surface vessel based on data-driven genetic algorithm", SHANDONG PROVINCIAL DEPARTMENT OF EDUCA-TION PROJECT (NO.J16LB72), SHANDONG KEY RESEARCH AND DEVELOPMENT PLAN (NO.2018YFJH0704), vol. 1, no. 2, 1 June 2019 (2019-06-01), pages 1 - 10, XP055786876, DOI: 10.11959/j.issn.2096−6652.201926 *

Cited By (55)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113191086A (zh) * 2021-05-08 2021-07-30 国网山东省电力公司营销服务中心(计量中心) 一种基于遗传算法的电采暖热负荷需求优化方法及系统
CN113517034A (zh) * 2021-05-21 2021-10-19 华东理工大学 分子筛工艺条件的生成方法和系统
CN113630737A (zh) * 2021-08-04 2021-11-09 西安电子科技大学 一种无线可充电传感器网络中移动充电器的部署方法
CN113780745A (zh) * 2021-08-16 2021-12-10 华中科技大学 一种上门服务需求驱动的it人员调度方法和系统
CN113780745B (zh) * 2021-08-16 2024-05-14 华中科技大学 一种上门服务需求驱动的it人员调度方法和系统
CN113625767A (zh) * 2021-09-02 2021-11-09 大连海事大学 一种基于优选信息素灰狼算法的固定翼无人机集群协同路径规划方法
CN113673782A (zh) * 2021-09-06 2021-11-19 中导光电设备股份有限公司 多显微镜扫描拍照路径优化方法和装置
CN113673782B (zh) * 2021-09-06 2024-02-23 中导光电设备股份有限公司 多显微镜扫描拍照路径优化方法和装置
CN113759915A (zh) * 2021-09-08 2021-12-07 广州杰赛科技股份有限公司 一种agv小车路径规划方法、装置、设备及存储介质
CN113759915B (zh) * 2021-09-08 2023-09-15 广州杰赛科技股份有限公司 一种agv小车路径规划方法、装置、设备及存储介质
CN113741477A (zh) * 2021-09-14 2021-12-03 中国船舶重工集团公司第七0七研究所九江分部 一种欠驱动船舶靠泊路径规划方法
CN113778128A (zh) * 2021-09-18 2021-12-10 中国人民解放军国防科技大学 基于遗传算法的无人机侦察任务规划方法及装置
CN113778128B (zh) * 2021-09-18 2023-03-31 中国人民解放军国防科技大学 基于遗传算法的无人机侦察任务规划方法及装置
CN113901085A (zh) * 2021-09-30 2022-01-07 中远海运科技股份有限公司 一种船舶轨迹动态绘制方法及系统
CN113901085B (zh) * 2021-09-30 2024-03-15 中远海运科技股份有限公司 一种船舶轨迹动态绘制方法及系统
CN113867358A (zh) * 2021-10-13 2021-12-31 中国人民解放军陆军工程大学 多无人车协同遍历任务的智能路径规划方法
CN113867358B (zh) * 2021-10-13 2023-08-29 中国人民解放军陆军工程大学 多无人车协同遍历任务的智能路径规划方法
CN114158105A (zh) * 2021-11-08 2022-03-08 石河子大学 应用于综合能源系统的无线传感器网络的路由方法及装置
CN114035587B (zh) * 2021-11-24 2024-03-29 陕西欧卡电子智能科技有限公司 一种无人船集群多船协同的路径规划方法、装置及无人船
CN114035587A (zh) * 2021-11-24 2022-02-11 陕西欧卡电子智能科技有限公司 一种无人船集群多船协同的路径规划方法、装置及无人船
CN113985899A (zh) * 2021-11-25 2022-01-28 江苏科技大学 基于区间多目标优化的水下机器人全局路径规划方法
CN113985899B (zh) * 2021-11-25 2023-09-22 江苏科技大学 基于区间多目标优化的水下机器人全局路径规划方法
CN114282435B (zh) * 2021-12-16 2024-03-08 青岛科技大学 一种基于改进遗传算法的无人船航线规划方法
CN114282435A (zh) * 2021-12-16 2022-04-05 青岛科技大学 一种基于改进遗传算法的无人船航线规划方法
CN114611243B (zh) * 2022-03-14 2023-09-08 华中科技大学 盾构通用型管片点位选取模型构建方法及点位选取方法
CN114611243A (zh) * 2022-03-14 2022-06-10 华中科技大学 盾构通用型管片点位选取模型构建方法及点位选取方法
CN114545621B (zh) * 2022-03-15 2023-09-22 中国科学院光电技术研究所 一种基于遗传算法的变形反射镜结构优化方法
CN114545621A (zh) * 2022-03-15 2022-05-27 中国科学院光电技术研究所 一种基于遗传算法的变形反射镜结构优化方法
CN114818465B (zh) * 2022-03-16 2024-02-23 西北工业大学 一种基于nurbs曲面的低散射优化方法
CN114818465A (zh) * 2022-03-16 2022-07-29 西北工业大学 一种基于nurbs曲面的低散射优化方法
CN115079704A (zh) * 2022-08-01 2022-09-20 中国电信股份有限公司 路径规划方法、装置、存储介质及电子设备
CN115079704B (zh) * 2022-08-01 2024-02-27 中国电信股份有限公司 路径规划方法、装置、存储介质及电子设备
CN115712507A (zh) * 2022-11-29 2023-02-24 上海船舶运输科学研究所有限公司 一种计算船舶网关任务优先级的方法
CN115712507B (zh) * 2022-11-29 2023-11-14 上海船舶运输科学研究所有限公司 一种计算船舶网关任务优先级的方法
CN115857516B (zh) * 2023-03-02 2023-07-14 之江实验室 结合牛耕式运动与遗传算法的全覆盖路径规划方法和装置
CN115857516A (zh) * 2023-03-02 2023-03-28 之江实验室 结合牛耕式运动与遗传算法的全覆盖路径规划方法和装置
CN116449685A (zh) * 2023-04-23 2023-07-18 上海交通大学 一种快速跟踪精准控制的舰船燃机控制方法及装置
CN116449685B (zh) * 2023-04-23 2024-03-29 上海交通大学 一种快速跟踪精准控制的舰船燃机控制方法及装置
CN116678409A (zh) * 2023-04-24 2023-09-01 大连海事大学 一种基于改进灰狼算法的北极冰区船舶路径规划方法
CN116678409B (zh) * 2023-04-24 2024-03-12 大连海事大学 一种基于改进灰狼算法的北极冰区船舶路径规划方法
CN116402317A (zh) * 2023-06-05 2023-07-07 山东梧桐城市规划技术服务有限公司 一种城市环境大数据综合管理系统
CN116402317B (zh) * 2023-06-05 2023-10-03 山东梧桐城市规划技术服务有限公司 一种城市环境大数据综合管理系统
CN117034480B (zh) * 2023-07-07 2024-03-22 北京航空航天大学 基于参数化降维的压气机多型面一体化优化方法及平台
CN117034480A (zh) * 2023-07-07 2023-11-10 北京航空航天大学 基于参数化降维的压气机多型面一体化优化方法及平台
CN117196019A (zh) * 2023-07-27 2023-12-08 南京中禹智慧水利研究院有限公司 基于改进自适应遗传算法的新安江模型参数率定方法
CN117196019B (zh) * 2023-07-27 2024-03-08 南京中禹智慧水利研究院有限公司 基于改进自适应遗传算法的新安江模型参数率定方法
CN116707624B (zh) * 2023-08-09 2023-11-07 北京天航华创科技股份有限公司 一种多艇组网通导遥系统及方法
CN116707624A (zh) * 2023-08-09 2023-09-05 北京天航华创科技股份有限公司 一种多艇组网通导遥系统及方法
CN117252114B (zh) * 2023-11-17 2024-02-13 湖南华菱线缆股份有限公司 一种基于遗传算法的电缆耐扭转实验方法
CN117252114A (zh) * 2023-11-17 2023-12-19 湖南华菱线缆股份有限公司 一种基于遗传算法的电缆耐扭转实验方法
CN117455918B (zh) * 2023-12-25 2024-03-26 深圳市辉熙智能科技有限公司 基于图像分析的防堵送料方法及系统
CN117455918A (zh) * 2023-12-25 2024-01-26 深圳市辉熙智能科技有限公司 基于图像分析的防堵送料方法及系统
CN117806328A (zh) * 2023-12-28 2024-04-02 华中科技大学 一种基于基准标记的无人艇靠泊视觉引导控制方法及系统
CN117689677A (zh) * 2024-02-01 2024-03-12 山东大学日照研究院 一种砂轮磨损状态识别方法、系统、设备及介质
CN117689677B (zh) * 2024-02-01 2024-04-16 山东大学日照研究院 一种砂轮磨损状态识别方法、系统、设备及介质

Also Published As

Publication number Publication date
LU101989B1 (en) 2021-03-09
CN110516877A (zh) 2019-11-29

Similar Documents

Publication Publication Date Title
WO2021035911A1 (zh) 一种基于数据正反向驱动线性变参数遗传算法的无人艇路径规划方法及系统
CN109933067B (zh) 一种基于遗传算法和粒子群算法的无人艇避碰方法
WO2021042827A1 (zh) 多agv的路径规划方法及系统
Yan et al. Three-dimensional path planning for autonomous underwater vehicles based on a whale optimization algorithm
KR20210149023A (ko) 해상 선박의 최적 경로를 결정하기 위한 장치
Zheng et al. A Decision‐Making Method for Ship Collision Avoidance Based on Improved Cultural Particle Swarm
EP3593292A1 (en) Training action selection neural networks
Das et al. Co-operative control of a team of autonomous underwater vehicles in an obstacle-rich environment
Wu et al. Robot path planning based on artificial potential field with deterministic annealing
Zhang et al. A multi-objective path planning method for the wave glider in the complex marine environment
WO2023082850A1 (zh) 行人轨迹预测方法、装置及存储介质
Du et al. Autonomous vehicle velocity and steering control through nonlinear model predictive control scheme
Bai et al. USV path planning algorithm based on plant growth
CN115994666A (zh) 基于自适应遗传学习粒子群算法的无人机任务分配方法
Wang et al. Obstacle avoidance for environmentally-driven USVs based on deep reinforcement learning in large-scale uncertain environments
Zhang et al. A multi-strategy improved differential evolution algorithm for UAV 3D trajectory planning in complex mountainous environments
Sharma et al. A genetic algorithm based nonlinear guidance and control system for an uninhabited surface vehicle
Zuo et al. A Robot SLAM Improved by Quantum‐Behaved Particles Swarm Optimization
CN117419716A (zh) 无人机三维路径规划方法、系统、存储介质和电子设备
CN117648548A (zh) 基于离线-在线混合强化学习的智能决策方法和装置
CN116661479A (zh) 建筑巡检路径规划方法、设备和可读存储介质
Liu et al. TD3 Based Collision Free Motion Planning for Robot Navigation
Song et al. A novel path planning algorithm for ships in dynamic current environments
CN113050658B (zh) 一种基于狮群算法优化的slam算法
Gong et al. A path planning method based on improved particle swarm optimization algorithm

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19943593

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19943593

Country of ref document: EP

Kind code of ref document: A1