A mobile robot time-varying path planning method based on code analysis template

Abstract

The application discloses a kind of mobile robot time-varying path planning methods based on encoding analysis template, it is related to path planning technical field, comprising: according to traffic network historical information and the data such as departure interval, construct time-speed ladder type function in specific space;Based on target distance, the formula of establishment of mobile robot arrival time in time period based on data such as departure time;With the minimum assignment cost, energy consumption cost and service time deviation as optimization goal, establish travel path model considering time-varying speed;Comprehensively consider service time window and other requirements, design initial solution generation algorithm based on target priority value;Using the improved genetic algorithm based on encoding analysis template, complete the joint optimization of mobile robot travel time and path planning;The present application is based on real scene, quickly plan the overall journey scheme of mobile robot, while meeting target demand, reduce comprehensive cost, provide more accurate decision support for city logistics distribution.

Description

Technical Field

[0001] This invention relates to the field of path planning technology, specifically a time-varying path planning method for mobile robots based on an encoding parsing template. Background Technology

[0002] As modern cities develop strategically around the integration of digital economy, digital life, and digital regulation, they are driving the integration of artificial intelligence with digital platform technologies such as 5G / 6G, mobile robots, cloud computing, blockchain, and the industrial internet, forming a modern "urban neural network system." In various scenarios such as CBDs, industrial parks, scenic areas, campuses, and communities, mobile robots are gradually replacing traditional logistics and delivery industries, building a new lifeline for digital cities.

[0003] Vehicle routing problem (VRP-TS) considering time-varying speed, as a technical field closer to actual delivery scenarios, has become one of the cutting-edge research areas in VRP in recent years. Time-varying speed refers to the treatment of vehicle speed as a function related to time. It can be integrated with logistics virtual systems through data acquisition, remote coordination, and real-time monitoring to complete simulation verification and dynamic adjustment under virtual-real integration, providing more accurate optimization analysis and decision support for the digital twin of urban logistics. To this end, this invention proposes a time-varying path planning method for mobile robots based on an encoded parsing template. Summary of the Invention

[0004] This invention aims to at least solve one of the technical problems existing in the prior art. To this end, this invention proposes a time-varying path planning method for mobile robots based on an encoded parsing template.

[0005] To achieve the above objectives, according to an embodiment of the first aspect of the present invention, a time-varying path planning method for a mobile robot based on an encoded parsing template is proposed, comprising the following steps:

[0006] Step 1: Based on historical traffic network information and departure interval data, construct a time-velocity step function in a specific space;

[0007] Step 2: Based on the target distance and departure time data, establish a formula for the arrival time of the mobile robot in different time periods;

[0008] Step 3: With the optimization objectives of minimizing assignment cost, energy consumption cost and service time deviation, establish a mobile robot travel path model that considers time-varying speed;

[0009] Step 4: Taking into account the service time window requirements, design an initial solution generation algorithm based on the target priority value; introduce the dual concepts of target urgency and target priority;

[0010] Step 5: Propose a new chromosome coding method: introduce a new type of coding method, namely coding analysis template; verify the effectiveness of the coding analysis template in the compression solution space through model input and output;

[0011] Step 6: Joint optimization of departure time and transportation path is realized by using improved genetic algorithm, and the optimal departure time and optimal path are output.

[0012] Further, the time-speed step function established in step 1 is represented as:

[0013]

[0014] Let vehicle k pass through road section (i, j), which spans p time intervals. l,τ , represents the length of time the vehicle travels in any time interval (τ = 1, 2,..., p+1); , represents the average travel speed of the vehicle in the T τ time interval; , represents the time when the vehicle departs from node i, , represents the time when the vehicle arrives at node j; d ij , represents the total distance of the (i, j) path; by solving a system of linear equations, the different travel speeds of the vehicle in the p time intervals are calculated

[0015]

[0016] Further, the time period mobile robot arrival time formula is established in step 2, which is specifically:

[0017] If the mobile robot arrives at the target node j in the current time interval τ: the vehicle arrival time If the mobile robot passes through two or more time intervals: the remaining distance Calculate whether it can arrive at j in the (τ+1) interval; loop to the time interval (τ+h) when it arrives, and output the arrival time

[0018] , the total distance of the (i, j) section d ij , the time when the vehicle departs from i , the time interval T τ = [b τ , e τ ] to which the departure time belongs, and the travel speed in that time zone

[0019] Further, the specific construction steps of the travel path model in step 3 are:

[0020]

[0021] Node access constraint

[0022]

[0023]

[0024] Route constraint in a certain time interval

[0025]

[0026]

[0027]

[0028] Target location load constraint

[0029]

[0030]

[0031] Time constraint

[0032]

[0033]

[0034]

[0035]

[0036]

[0037]

[0038]

[0039]

[0040]

[0041] Sub-loop constraint

[0042]

[0043]

[0044]

[0045] Wherein O represents a distribution center; C represents a target location set, C = {1, 2,..., n C};V represents all nodes, V = {0} U C; K represents a set of mobile robot vehicles, K = {1, 2,..., n K};u i represents the loading capacity of the vehicle at the target site node i; q ik represents the loading capacity of the vehicle k when leaving the node i; d ij represents the distance between the point i and the point j; Q represents the maximum vehicle loading capacity; [E i ,L i ] represents the service time window of the node i; s i represents the service time of the target site i; represents the time when the vehicle k arrives at the point i; represents the time when the vehicle k leaves the point i; represents the time when the service of the target site i starts; T represents a set of time intervals, T = {1, 2,..., n t};[b τ ,e τ ] represents the start and end time of the time interval τ; represents the time when the vehicle k passes through the arc (i, j) in τ; represents the distance when the vehicle k passes through the arc (i, j) in τ; w1 represents the unit vehicle use cost; w2 represents the unit energy consumption cost; w3 represents the unit time deviation cost; M is a large enough number; represents when the vehicle k travels from the node i to the node j, is 1, otherwise 0; represents when the vehicle k travels from the node i to the node j in the time interval τ, is 1, otherwise 0.

[0046] Further, an initial solution generation algorithm based on the target priority value is designed in step 4, specifically:

[0047] target urgency u j is the latest time when the mobile robot starts to serve the target site, which is expressed as the latest service end time acceptable to the target minus the service time;

[0048] target priority value c ij is defined as: the weighted integration of the target site time window offset and the travel time from the current node to the target site under the unified departure time; the smaller the target priority value is, the shorter the remaining time for departure to the target site node is, and the mobile robot prefers to go to;

[0049] u j = L j -s j

[0050]

[0051] c ij = alpha1 h i + alpha2 t ij , alpha1 + alpha2 = 1

[0052] The mobile robot departs from the distribution center, and a target location node with a small target priority value is preferentially selected and inserted into the current path, and the insertion is repeated until there is no target location that can be served in the current path; if there is still a target location that has not been served, an empty path is constructed, and the above process is repeated.

[0053] Further, the encoding analysis template in step 5 is represented as:

[0054] n target locations and k vehicles are coded, and f(n, k) chromosomes are designed in total; the number of genes on each chromosome is the total number of vehicles k, the cumulative sum of the genes is the number of target locations n, and the genes are positive integers not less than 1; the mathematical expression is n = n1 + n2 + … + nk, and n1 >= n2 >= … >= nk >= 1. k k

[0055] Wherein,

[0056] Further, the joint optimization of the departure time and the transportation path in step 6 is as follows:

[0057] The earliest initial feasible departure time is selected from the feasible set of vehicle departure times;

[0058] The optimal path is determined according to the departure time by using a genetic algorithm, and the next time is selected for repeated operation until the retrieval of all feasible times is completed;

[0059] Based on the minimum target cost, the optimal travel time and optimal path information are output.

[0060] Compared with the prior art, the beneficial effects of the present application are:

[0061] The present application constructs a time-speed ladder function in a specific space according to historical information of the traffic network and data such as departure intervals. A travel path model considering time-varying speed is established with the optimization objectives of minimizing the assignment cost, energy cost and service time deviation, and an improved genetic algorithm based on the encoding analysis template is used to complete the joint optimization of the travel time and path planning of the mobile robot. The present application relies on real scenarios to quickly plan the overall travel scheme of the mobile robot, meets the target demand while reducing the comprehensive cost, and provides more accurate decision support for urban logistics. BRIEF DESCRIPTION OF DRAWINGS

[0062] ​​In order to make the technical solutions of the embodiments of the present application or the prior art clearer, the accompanying drawings needed in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description only need to be some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort.

[0063] Fig. 1 The figure is a schematic diagram of the principle of the present application.

[0064] Fig. 2 The figure is a flow chart of the algorithm for determining the optimal path by using the genetic algorithm in the present application. DETAILED DESCRIPTION

[0065] The technical solutions of the present application will be described clearly and completely in combination with the embodiments. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort are within the protection scope of the present application.

[0066] As shown in Figs. 1-2 , a time-varying path planning method for mobile robots based on coding analysis template includes the following steps:

[0067] Step 1: According to the historical information of the traffic network and the data such as the departure interval, a time-velocity step function in a specific space is constructed. Specifically:

[0068] When the mobile robot vehicle travels on a road section, it may cross one or more time intervals, and the average travel speed changes. Suppose that the vehicle k passes through the road section (i, j) and travels across p time intervals. In the T τ (τ=1,2,...,p+1) time interval, the average travel speed of the vehicle is The time when the vehicle starts from node i is The time when the vehicle arrives at node j is The specific implementation steps for obtaining the related parameters of the p+1 time interval are as follows:

[0069] The travel speed of the mobile robot in the time period is calculated by inputting the speed state parameters. The time length of the vehicle traveling in any time interval T τ (τ=1,2,...,p+1) is:

[0070]

[0071] The travel speed of the vehicle in different time intervals The travel time t l,τ and the total distance d of the pathij The relationship is expressed in a linear equation set, and the different driving speeds of the vehicle in the p time intervals are calculated by solving the equation set

[0072]

[0073] Step 2: Based on the target distance, departure time and other data, the moving robot arrival time formula is established; specifically:

[0074] The moving robot arrives at the target location node j in the current time interval τ: vehicle arrival time

[0075] The moving robot passes through two or more time intervals: remaining distance Calculate whether it can arrive at j in the (τ+1) interval. Loop to the time interval (τ+h) where arrival is achieved, and output the arrival time

[0076] Where, the total distance d of the (i,j) section ij The time when the vehicle departs from i The load q at the time of departure ik The time interval T to which the departure time belongs τ =[b τ ,e τ ] and the driving speed in this time interval

[0077] Step 3: Establish a moving robot travel path model considering time-varying speed, with the optimization goal of minimizing assignment cost, energy consumption cost and service time deviation; specifically including:

[0078]

[0079] Node access constraints

[0080]

[0081]

[0082] Route constraints in a certain time interval

[0083]

[0084]

[0085]

[0086] Target location load constraints

[0087]

[0088]

[0089] Time constraints

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099] Sub-loop constraints

[0100]

[0101]

[0102]

[0103] The model is represented as follows: 0 represents the distribution center; C represents the target point set, C = {1, 2,..., n C}; V represents all nodes, V = {0}∪C; K represents the mobile robot vehicle set, K = {1, 2,..., n K}; u i represents the loading amount of the vehicle at the target point node i; q ik represents the load of the vehicle k when leaving the node i; d ij represents the distance between points i and j; Q represents the maximum vehicle load capacity; [E i ,L i ] represents the service time window of node i; s i represents the service time of the target point i; represents the time when the vehicle k arrives at point i; represents the time when the vehicle k leaves point i; represents the time when the service to the target point i starts; T represents the time interval set, T = {1, 2,..., n T}; [b τ ,eτ denotes the start and end time of time interval τ; denotes the time vehicle k travels through arc (i, j) in τ; denotes the distance vehicle k travels through arc (i, j) in τ; w1 denotes the unit vehicle usage cost; w2 denotes the unit energy cost; w3 denotes the unit time deviation cost; M is a large enough number; denotes when vehicle k travels from node i to node j, is 1, otherwise 0; denotes when vehicle k travels from node i to node j in time interval τ, is 1, otherwise 0.

[0104] wherein (1) is the optimization objective, i.e. minimization of operation cost (departure cost and energy cost) and service time deviation. (2) denotes that each destination can only be served once by each vehicle. (3) denotes the in-out balance constraint, i.e. vehicle must enter and exit from a node. (4) denotes that vehicle must travel through arc (i, j) in one or more time intervals when transporting from node i to node j. (5) denotes that the distance vehicle travels through arc (i, j) in a time interval τ is no more than the total distance of arc (i, j). (6) denotes that the sum of the distance vehicle travels through arc (i, j) in all time intervals is equal to the total distance of arc (i, j). (7) denotes the change of vehicle load after completing loading and unloading at the front and rear nodes. (8) denotes that the load of vehicle when leaving a node must not exceed the maximum vehicle load. (9) denotes that the travel distance of vehicle in a time interval is proportional to the travel speed. (10) denotes that the travel time of vehicle in a time interval τ is no more than the unit time interval period. (11) denotes the relationship between the time vehicle leaves node i and the travel time of vehicle in time interval τ. (12) denotes the relationship between the time vehicle arrives at node j and the travel time of vehicle in time interval τ. (13) denotes that the time vehicle leaves node i plus the travel time of vehicle is greater than or equal to the time vehicle arrives at node j. (14) denotes that the time vehicle starts serving destination i should be within the time window of i. (15) denotes that the service time of vehicle is the larger value between the arrival time of vehicle and the starting time window of destination. (16) denotes the service time constraint of vehicle at destination. (17) denotes that vehicle should return to the distribution center before the latest working time of the distribution center. (18) denotes the removal of sub-loop constraint. (19)-(20) are 0-1 decision variables.

[0105] Step 4: design an initial solution generation algorithm based on target priority value considering the requirements such as service time window; specifically including:

[0106] When the algorithm generates an initial solution, a random assignment or the earliest completion time or the nearest target node insertion into the current path can be selected in general. To improve the quality of the initial solution, the invention introduces the concepts of target urgency and target priority. The target urgency u j The time at which the mobile robot starts to serve the target location is expressed as the latest service end time acceptable to the target - service time. The target priority c ij is defined as the weighted integration of the target location time window offset and the travel time from the current node to the target location at the uniform departure time. The smaller the target priority, the shorter the remaining time for the departure of the mobile robot to the target location, and the mobile robot prefers to go to it.

[0107] u j = L j -s j

[0108]

[0109] c ij = α1h i + α2t ij , α1+ α2 = 1

[0110] The basic idea of the initial solution algorithm based on target priority (ECP) is that the mobile robot departs from the distribution center, and preferentially selects a target location node with a small target priority indicator to insert into the current path, and repeats the insertion until there is no target location that can be served on the path. If there are still unserved target locations, an empty path is constructed, and the above process is repeated.

[0111] Step 5: A new chromosome coding method is proposed: coding analysis template, verify its effectiveness in compressing the solution space; specific includes:

[0112] A new type of coding method, coding analysis template, is considered to be introduced. Coding is performed on n target locations and k vehicles, and a total of f(n, k) chromosomes are designed. The number of genes on each chromosome is the total number of vehicles k, the cumulative sum of each gene is the number of target locations n, and the gene is a positive integer not less than 1. The mathematical expression is n = n1+ n2+…+ nk k and n1≥ n2≥…≥ nk≥ 1; k

[0113] wherein,

[0114] ​The effectiveness of the encoding analysis template is verified by model input and output. By encoding grouping, the number of times of using the method is low and the solution space can be greatly reduced. The number of target locations n = 8 and the number of vehicles k = 2 are known, that is, 8-bit target locations can be divided into two groups, and each group has no less than 1-bit target location. Without considering the repetition, when a group of target location genes is randomly generated, four grouping analysis results are obtained by the above formula, and the genes can be substituted into four templates at a time; at this time, all vehicle grouping modes have been accurately fixed, that is, (7+1), (6+2), (5+3), (4+4), and there is no other effective grouping.

[0115] In this example, the solution space of this encoding mode contains 4*n! different solutions in total, and the redundant solutions caused by vehicle index, path loop, etc. are excluded. As shown in Table 1, compared with three commonly used encoding modes, the solution space of the encoding analysis template is the smallest, and the operation efficiency is significantly improved.

[0116] Table 1 Comparison of encoding mode effects

[0117]

[0118] The template effect is tested under different target location scales. It is assumed that the number of mobile robot cars k = 10 is given, and the number of target locations n is 30, 50, and 100 respectively, and the corresponding number of encoding analysis templates is shown in Table 2:

[0119] Table 2

[0120] n f(n, k) 30 530 50 16928 100 2977866

[0121] The solution space values of the four encoding modes are compared in Table 3, wherein the solution space of the encoding analysis template is the smallest in various target location conditions, verifying the effectiveness of the method.

[0122] Table 3

[0123]

[0124]

[0125] Step 6: The improved genetic algorithm is used to realize joint optimization of the departure time and the transportation path, and the optimal departure time and the optimal path are output; specifically including:

[0126] When the speed of the mobile robot changes with time, the vehicles will have different speeds when they start at different time points, and the energy consumption and logistics cost related to the speed will also change accordingly, so the present application considers optimizing the departure time of the vehicles while optimizing the transportation path.

[0127] For example Fig. 2The main solution method is to select the earliest initial feasible departure time from the feasible set of vehicle departure time; the optimal path is determined by genetic algorithm according to the departure time, and the next time is selected for repeated operation until the retrieval of all feasible time is completed. Based on the minimum transportation cost, the optimal travel time and optimal path information are output.

[0128] The above formulas are dimensionless values calculated, and the formulas are obtained by software simulation of a large amount of data to obtain a formula closest to the actual situation. The preset parameters and preset thresholds in the formula are set by a person skilled in the art according to the actual situation or obtained by a large amount of data simulation.

[0129] In the description of the present specification, the description of the terms "one embodiment", "example", "specific example" and the like means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0130] The preferred embodiments of the application disclosed above are only used to help explain the application. The preferred embodiments do not describe all the details and limit the application to the specific embodiments. Obviously, many modifications and changes can be made according to the content of the present specification. The present specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the application, so that those skilled in the art can well understand and utilize the application. The application is limited only by the claims and their full scope and equivalents.

Claims

1. A time-varying path planning method for mobile robots based on encoding resolution templates, characterized in that, The method comprises the following steps: Step 1: constructing a time-speed step function in a specific space according to historical information of a traffic network and departure interval data; wherein the time-speed step function is expressed as: Supposing that the vehicle k passes through the road section (i, j) and travels across p time intervals; t l，τ denotes the length of time the vehicle travels in any time interval τ = 1, 2,..., p + 1; denotes the time of arrival at node j; denotes the start and end times of time interval τ; Step 2: establishing a formula for the arrival time of the mobile robot in different time periods based on target distance and departure time data; Step 3: establishing a mobile robot travel path model considering time-varying speed, with the optimization objective of minimizing assignment cost, energy consumption cost and service time deviation; the specific construction steps of the travel path model are as follows: where: o represents a distribution center; C represents a target point set, C = {1, 2,..., n C}; V represents all nodes, ; K represents a mobile robot vehicle set, K = {1, 2,..., n K}; w1 represents a unit vehicle use cost; w2 represents a unit energy consumption cost; w3 represents a unit time deviation cost; represents a time when vehicle k arrives at point i; represents a service time window of node i; represents when vehicle k travels from node i to node j, 1, otherwise 0; Step 4: designing an initial solution generation algorithm based on target priority value, considering the requirement of service time window; introducing the concepts of target urgency and target priority; Step 5: constructing a new chromosome coding mode, i.e., a coding analysis template; verifying the effectiveness of the coding analysis template in compressing the solution space through model input and output; the coding analysis template is expressed as: For n target locations and n K The vehicles are coded, and a total of f(n,n) is designed. K Chromosomes; the number of genes on each chromosome is equal to the total number of vehicles, n. K The sum of all genes equals the number of target locations, n, and the genes are positive integers not less than 1; the mathematical expression is: ; wherein ; Step 6: realizing joint optimization of the departure time and the transport path by using an improved genetic algorithm, and outputting the optimal departure time and the optimal path.

2. The time-varying path planning method for mobile robots based on coding and parsing templates according to claim 1, characterized in that, The step 1 further comprises: solving a system of linear equations to calculate different driving speeds of the vehicle over the p time intervals : ； where d ij denotes the distance between point i and point j.

3. The time-varying path planning method for mobile robots based on coding and parsing templates according to claim 2, characterized in that, The step 2 of establishing the formula for the arrival time of the mobile robot in different time periods comprises the following specific steps: If the mobile robot reaches the target location j in the current time interval τ: vehicle arrival time ; If the mobile robot passes two or more time intervals: remaining distance ; calculate whether it can reach j in τ+1 interval; loop to the time interval τ+h in which it achieves the arrival, and output the time of arrival at node j ; wherein the time of departure of the vehicle from node i the time interval to which the time of departure belongs and the travel speed within this time interval .

4. The time-varying path planning method for mobile robots based on coding and parsing templates according to claim 3, characterized in that, The travel path model further comprises: Node access constraint wherein, Road distance constraint in a certain time interval Target site load constraint Time constraint Sub-loop constraint wherein, u i denotes the loading amount of the vehicle at the target location i; q ik denotes the loading amount of the vehicle at the target location i; q i denotes the loading amount of the vehicle at the target location i; q i denotes the service time window of the node i; s i denotes the service time of the target location i; denotes the time when the service to the target location i starts; T denotes a set of time intervals, T = {1, 2,..., n T}; denotes the time when the vehicle departs from the node i; ; denotes the time when the vehicle k passes through the arc (i, j) in the time interval τ; denotes the distance when the vehicle k passes through the arc (i, j) in the time interval τ; denotes the distance when the vehicle k passes through the arc (i, j) in the time interval τ; is 1, otherwise 0.

5. The time-varying path planning method for mobile robots based on coding and parsing templates according to claim 2, characterized in that, The step 4 of designing the initial solution generation algorithm based on the target priority value comprises the following specific steps: Target urgency u j The time at which the mobile robot must start serving the target location for the last time, expressed as the target-acceptable latest end-of-service time - service time. Target priority value c ij Definition: the weighted integration of the target location time window offset and the travel time from the current node to the target location under the unified departure time; the smaller the target priority value, the shorter the remaining time for the departure to the target location, and the mobile robot prefers to go there; The mobile robot departs from the distribution center, and preferentially selects a target site with a small target priority value to be inserted into the current path, and repeats the insertion until there is no target site to be served in the current path; if there is still an unserved target site, an empty path is constructed, and the above process is repeated.

6. The time-varying path planning method for mobile robots based on coding and parsing templates according to claim 1, characterized in that, The joint optimization of the departure time and the transport path in step 6 is specifically as follows: An initial feasible departure time is selected from the feasible set of vehicle departure times; The optimal path is determined by using the genetic algorithm according to the departure time, and the next time is selected for repeated operation until the retrieval of all feasible times is completed; Based on the minimum transport cost, the optimal travel time and the optimal path information are output.

Images