Mobile robot trajectory planning method and device based on two-step optimization strategy

Abstract

The invention relates to a mobile robot trajectory planning method based on a two-step optimization strategy, and the method comprises the following steps: employing a maximum inscribed circle to represent the collision range of a mobile robot, simplifying the modeling of collision constraints, employing control points of a spline line as optimization variables, and enabling the number of the control points to be adjusted in a self-adaptive manner along with the environment, a polynomial curve obtained through optimization is used as an initial value of the next step of optimization; the chassis information of the mobile robot is accurately described by using an Euclidean distance field, adjacently connected path points in a cubic polynomial curve are used as an optimization objective function, and an optimization pose sequence and a path curve of the robot are solved through an algorithm. The invention also relates to a computer device comprising a memory and a processor, wherein the processor implements the trajectory planning method when executing instructions stored in the memory.

Description

Technical Field

[0001] The present invention relates to a mobile robot trajectory planning technical solution, in particular to a mobile robot trajectory planning method and device based on a two-step optimization strategy, and belongs to the technical field of robot automatic control. Background Art

[0002] Mobile robots have broad applications and development prospects in home services, factory logistics, etc. However, complex environments place higher demands on mobile robot planning and navigation. In extremely narrow environments, such as doors, corridors and other narrow channel environments, navigation is still a difficult problem that mobile robots need to solve. Traditional path sampling methods cannot complete navigation in such environments due to limited sampling resolution. General optimization methods often have difficulty in solving collision-free navigation trajectories due to unreasonable constraint modeling and initial value settings, and cannot achieve stable navigation in narrow spaces.

[0003] In addition, real-time online planning and navigation of mobile robots in confined scenes is even more difficult, which places higher requirements on the accuracy and real-time performance of the planning algorithm solution. In traditional trajectory planning optimization methods, due to unreasonable selection of initial values, more computing resources will be used to iterate and obtain a suitable trajectory line, thereby increasing the consumption of onboard computing platform resources and increasing computing time, affecting the real-time performance of planning. Summary of the invention

[0004] The present invention provides a mobile robot trajectory planning technical solution, aiming to solve at least one of the technical problems existing in the prior art, such as solving the technical problem that the current navigation planning algorithm occupies a large amount of computing resources and is time-consuming to calculate in a narrow channel environment.

[0005] In some aspects, the technical solution of the present invention is a path planning method for a mobile robot, comprising the following steps:

[0006] S100, acquiring geometric data of a maximum inscribed circle according to the chassis shape of the mobile robot, and configuring a set of control points of a preset number at an initial sampling resolution to be associated with a trajectory line of path planning of the mobile robot;

[0007] S200, updating the collision data of the mobile robot, constructing a collision avoidance constraint, and optimizing the positions of a set of control points based on a first objective function of minimum path length consumption under a collision-free condition, and fitting the control points obtained by the optimization calculation into an updated trajectory line;

[0008] S300, determining that the maximum inscribed circle of the mobile robot does not collide with obstacles when moving along all strokes of the updated trajectory line, then obtaining the polynomial curve trajectory geometry data under the inscribed circle constraint model as the initial value of the refined optimized trajectory, otherwise, adding more control points in the collision area, and returning to step S200;

[0009] S400, establishing a Euclidean distance field model of the mobile robot according to the precise geometric shape data obtained from the chassis shape of the mobile robot, adding a collision risk area, and constructing a collision constraint;

[0010] S500, using the curve generated by the polynomial curve trajectory geometry data as the connecting line of the path points of the mobile robot, establishing the posture sequence of the mobile robot in the path points as the optimization variable, and solving the optimized posture sequence based on the optimization equation of the second objective function of minimum energy consumption, and fitting the posture sequence obtained by the optimization solution into an updated polynomial curve;

[0011] S600, determine that the precise geometric shape of the mobile robot does not collide with obstacles along all the strokes of the fitted polynomial curve, then obtain the coefficient parameters of the polynomial curve and the optimized posture sequence as the path planning result of the mobile robot; otherwise, increase the posture sampling density in the collision area to generate a posture sequence, and return to step S400.

[0012] Further, the step S100 includes:

[0013] Set the risk area and provide the risk penalty value calculation function f through the following formula obs , used for the optimization calculation of the first objective function of minimum energy consumption,

[0014]

[0015] Where, ρ represents the distance between the center of the chassis of the mobile robot and the edge of the obstacle, r m represents the maximum inscribed circle radius of the chassis shape of the mobile robot, c represents the maximum value of the risk area, δ represents the maximum distance from the edge of the risk area to the edge where the collision occurs,

[0016] If <r m Then the mobile robot is in the collision area if r m < <r m +δ then the mobile robot is in the risk area.

[0017] Furthermore, the step S100 further includes:

[0018] Preliminary curve sampling is performed along the paths of the mobile robot's movement starting point, movement target point and obstacle detour points to evenly configure a group of control points, the initial configuration number of which is substantially equal to the total number of points of obstacle detour, movement starting point and movement target point.

[0019] Further, the step S200 includes the following steps:

[0020] S210, the first objective function with minimum energy consumption is obtained by the following formula The gradient descent algorithm is used to optimally solve the position vector q of the control point.

[0021]

[0022]

[0023] Among them, J b represents the collision constraint, J sm Indicates the drastic change of the control point, J le represents the length constraint of the trajectory line, λ b , sm , le They represent the constraint coefficient, smoothness coefficient and path length coefficient of collision avoidance, respectively. obs represents the risk penalty value calculation function, o j Represents the obstacle position vector. i, j, n, and m are natural numbers and are used for subscript and subscript counting of parameters in the formula.

[0024] Furthermore, the step S200 further includes the following steps:

[0025] S220, fitting the control points obtained by optimization calculation into a spline curve p by the following formula i (τ) and as the trajectory,

[0026]

[0027] Among them, τ is the function variable used to generate the spline curve, τ∈[0,1],n p is the number of path points of the mobile robot, The constant matrix representing the spline curve, s(τ) is the vector of polynomial terms that constitute the spline curve, π i is the control point matrix, i is a natural number used to count the subscripts and subscripts of the parameters in the formula, i≤n q -n p , n q is the number of control points, which must be no less than the number of path points.

[0028] Further, the step S400 includes:

[0029] Based on the Euclidean distance field model, the collision risk value calculation function is provided by the following formula To establish collision constraints, solve the optimization equation of the second objective function for minimum energy consumption,

[0030]

[0031] Among them, o represents the obstacle vector, c represents the maximum value of the risk area, represents the Euclidean distance field model operation, represents the shape space of the mobile robot chassis, represents the expanded shape space of the mobile robot chassis after superimposing the risk area, Indicates that the obstacle is within the collision range. Indicates that the obstacle is within the risk area, c represents the maximum value of the risk area, δ represents the maximum distance from the edge of the risk area to the edge where the collision occurs, and min{||op||} represents the shortest distance from the obstacle to the boundary.

[0032] Further, the step S500 includes the following steps:

[0033] S510, the second objective function with minimum energy consumption is calculated by the following formula Optimally solve the posture sequence z of the mobile robot,

[0034]

[0035] J end =λ end ||z i -z e ||

[0036]

[0037] Among them, J smo represents the smoothness constraint, j end represents the terminal pose constraint, J kin Indicates arc constraint, J v represents no collision constraint, λ smo represents the smoothness penalty coefficient matrix, λ end represents the penalty coefficient of the optimization objective, λ kin represents the arc constraint coefficient, λ v represents the non-collision constraint coefficient, δ represents the maximum distance from the edge of the risk area to the edge where the collision occurs, h represents the kinematic constraint function of the mobile robot, and z i represents the pose vector, z e represents the end pose vector, represents the collision risk value calculation function, Ri represents the posture matrix of the mobile robot, o j represents the position vector of the obstacle, p i Represents the position vector of the mobile robot. i and j are natural numbers and are used for subscript and subscript counting of parameters in the formula.

[0038] Further, the step S500 includes the following steps:

[0039] S520, performing cubic polynomial curve fitting in the X and Y directions on the posture sequence z obtained by optimization using the following formula:

[0040]

[0041] Where τ is the function variable used to generate the cubic polynomial curve, s(τ) is the vector of polynomial terms that constitute the cubic polynomial curve, and d a and d b are the parameter matrices of the polynomial terms in the X and Y directions respectively;

[0042] Among them, the constraint condition AX is given according to the requirement that the starting and ending points of the trajectory line of adjacent posture points in the posture sequence of the mobile robot need to meet the requirement that the position and posture coincide. k = 0, and the energy optimization is used as the objective function through the following formula Optimally solve the optimization variable X of the cubic polynomial curve k ,

[0043]

[0044] sxt k =0

[0045] in,

[0046]

[0047] η k and γ k is the relaxation factor coefficient, B and G are constant matrices, and k is the formula letter subscript.

[0048] The technical solution of the present invention also relates to a computer device, comprising a memory and a processor, wherein the processor implements the above method when executing a computer program stored in the memory.

[0049] The technical solution of the present invention also relates to a mobile robot, comprising: an environmental sensor for obtaining position and shape data of obstacles; a posture sensor for feeding back position and posture data of the mobile robot; and the above-mentioned computing device, which is respectively connected to the environmental sensor and the posture sensor.

[0050] Therefore, the scheme proposed by the present invention mainly involves two steps of optimization: the first step is to use the maximum inscribed circle to represent the collision range of the mobile robot, simplify the modeling of the collision constraint, and use the control points of the cubic spline as the optimization variable, and the control points can be adaptively adjusted with the environment. The polynomial curve obtained after optimization is used as the initial value of the next step of optimization; the second step is to use the Euclidean distance field to accurately describe the chassis information of the mobile robot, and use the adjacent connected path points in the cubic polynomial curve as the optimization objective function, and solve it through the LBFGS (Limited-memory Broyden Fletcher Goldfarb Shanno) algorithm. The beneficial effect is that in the passage of the narrow space, the rough collision modeling is used in the first step to expand the solution space, which can avoid considering the optimization modeling of the attitude angle, greatly reduce the calculation time, and can obtain a better initial value for the second step process; in the second step, an accurate mobile robot collision model can be established, so that the planning result is better and more accurate.

[0051] In addition, additional aspects and advantages of the present invention will be given in part in the following description, and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 is a flowchart of a planning method according to an embodiment of the present invention.

[0053] Figure 2 It is the theoretical shape of the mobile robot and its largest inscribed circle, and a schematic diagram representing its relationship with obstacles.

[0054] Figure 3 It is a schematic diagram representing the impact of risk areas on mobile robot planning.

[0055] FIG4 is a schematic diagram showing the influence of control points of different resolutions on planning.

[0056] Figure 5 It is a schematic diagram of inserting control points between the control points where collision occurs.

[0057] Figure 6 It is the numerical value of the Euclidean distance field of the mobile robot chassis and its gradient diagram.

[0058] Figure 7 It is a schematic diagram of obstacle avoidance control point planning with a sampling interval of 0.5m in an example according to the present invention.

[0059] Figure 8 It is a schematic diagram of obstacle avoidance planning with adaptive control point adjustment at a sampling interval of 0.5m in an example of the present invention.

[0060] Fig. 9It is a planning diagram in a random obstacle environment.

[0061] Fig.10 It is a schematic diagram of planning in a channel environment with a width of 0.7m in an example according to the present invention.

[0062] Fig.11 It is a schematic diagram of planning in a narrow curved channel in an example according to the present invention.

[0063] Fig.12 1 is a diagram of a physical test scene of a mobile robot in an example according to the present invention. DETAILED DESCRIPTION

[0064] The concept, specific structure and technical effects of the present invention will be clearly and completely described below in combination with the embodiments and drawings to fully understand the purpose, scheme and effect of the present invention.

[0065] Reference Figure 1 In some embodiments, the method according to the present invention basically comprises the following steps:

[0066] S100, acquiring geometric data of a maximum inscribed circle according to the chassis shape of the mobile robot, and configuring a set of control points of a preset number at an initial sampling resolution to be associated with a trajectory line of path planning of the mobile robot;

[0067] S200, updating the collision data of the mobile robot, constructing collision avoidance constraints, and optimizing the positions of a set of control points based on a first objective function of minimum energy consumption, and fitting the control points obtained by the optimization calculation into an updated trajectory line;

[0068] S300, determining that the maximum inscribed circle of the mobile robot does not collide with obstacles when moving along all strokes of the updated trajectory line, then obtaining the polynomial curve trajectory geometry data under the inscribed circle constraint model as the initial value of the refined optimized trajectory, otherwise, adding more control points in the collision area, and returning to step S200;

[0069] S400, establishing a Euclidean distance field model of the mobile robot according to the precise geometric shape data obtained from the chassis shape of the mobile robot, adding a collision risk area, and constructing a collision constraint;

[0070] S500, using the curve generated by the polynomial curve trajectory geometry data as the connecting line of the path points of the mobile robot, establishing the posture sequence of the mobile robot in the path points as the optimization variable, and solving the optimized posture sequence based on the optimization equation of the second objective function of minimum energy consumption, and fitting the posture sequence obtained by the optimization solution into an updated polynomial curve;

[0071] S600, determine that the precise geometric shape of the mobile robot does not collide with obstacles along all the strokes of the fitted polynomial curve, then obtain the coefficient parameters of the polynomial curve and the optimized posture sequence as the path planning result of the mobile robot; otherwise, increase the posture sampling density in the collision area to generate a posture sequence, and return to step S400.

[0072] The implementation of the above steps is described in detail below.

[0073] Specific implementation of step S100

[0074] As attached Figure 2 As shown, the chassis of the mobile robot is a polygon that approximates a rectangle. First, the maximum inscribed circle of the rectangle is obtained, where the maximum radius of the inscribed circle is r m , and establish a risk area, δ represents the width of the risk area, c represents the maximum penalty value of the risk area, to ensure that the mobile robot keeps a certain distance from obstacles as much as possible under the premise of ensuring safety. If the passing area is relatively narrow, the planned trajectory will tend to the middle of the narrow space, and the obstacle may fall into the risk area, but will not fall into the inscribed circle. If the passing area is wide, the planned trajectory will be as far away from the obstacle as possible, so that the obstacle will not fall into the risk area. The established forbidden area and risk area function f obs (ρ) is as follows:

[0075]

[0076] Among them, if ρ <r m Then the mobile robot will collide with the obstacle, satisfying r m < <r m +δ, then the mobile robot is in the risk area and will not collide with obstacles. However, if the control accuracy of the mobile robot is insufficient, there is a risk of collision. Figure 3 The schematic diagram of the impact of risk areas on mobile robot planning is shown in Figure 1 (the dotted area in the figure is the prohibited area, and the expanded area outside the dotted line is the risk area). The mobile robot encounters r when passing through two obstacles. m < <r m +δ case.

[0077] In addition, as shown in Figure 4, using the control point sampling resolution in Figure 4 (a) will result in more optimization target variables, which will increase the calculation time; using the larger intervals in Figure 4 (b) to set the control point distribution will reduce the scale of the optimization problem and reduce the calculation time, but the trajectory between the control points may collide with obstacles. The initial control points use a larger interval to reduce the calculation time, and in step S300, the control points will be inserted adaptively to achieve a reasonable control point distribution, ensuring that the calculation time is short and safe without collision.

[0078] Specific implementation of step S200

[0079] Construct collision avoidance constraints, and use trajectory smoothness and shortest trajectory as the objective function to be solved to obtain the optimal control point position, where J b represents the collision constraint, J sm Indicates the drastic change of the control point, J le represents the length constraint of the trajectory line, λ b , sm , le They represent the constraint coefficient, smoothness coefficient, and path length coefficient of collision avoidance, respectively, as shown in the following formula:

[0080]

[0081]

[0082] And use the gradient descent algorithm to optimize and solve. The spline formed by the control points can be obtained as follows:

[0083]

[0084] where τ∈[0,1],i=0,1,…,n q -n p , The constant matrix representing the spline, the spline curve formed according to the above control points, is used as the initial value in the fine optimization process.

[0085] Specific implementation of step S300

[0086] It is necessary to perform a collision check on the spline calculated by the optimized control points to check whether the trajectory line obtained by the optimized calculation collides with the obstacle. If a collision occurs, a new control point is inserted between the colliding control points, such as Figure 5 As shown, and return to step S200 to calculate again. If no collision occurs, then the polynomial curve trajectory under the rough inscribed circle constraint model is obtained as the initial value of the fine optimization. For example, Figure 7 This is a schematic diagram of obstacle avoidance control point planning with a sampling interval of 0.5m. Figure 7 If the calculated trajectory line collides with an obstacle during the inspection, the interval adaptive control points need to be adjusted to generate a new obstacle avoidance planning curve. The effect obtained is as follows: Figure 8 shown.

[0087] The specific implementation of step S400 is as follows:

[0088] Establish an accurate Euclidean distance field (EDF) model for mobile robots. The Euclidean distance field can easily query the distance and gradient information of obstacles, which plays an important role in the online motion planning of mobile robots. Directly using the Euclidean distance field for the optimization process will cause the mobile robot to be too close to the obstacle. Therefore, a risk area is added to keep the obstacle as far away from the mobile robot as possible. If the obstacle falls outside the risk area, its collision penalty value is 0. The collision penalty function used for the constraint The build is as follows,

[0089]

[0090] If the obstacle is within the risk area, c represents the maximum value of the risk area, δ represents the maximum distance from the edge of the risk area to the edge where the collision occurs, so its value in the risk area is Where min{||op||} represents the shortest distance from the obstacle to the boundary. The specific field diagram is shown in Figure 6 .

[0091] Specific implementation of step S500

[0092] Taking the pose sequence as the optimization target variable, the optimization equation using smoothness as the optimization target function is

[0093]

[0094] Among them J smo The smoothness constraint is expressed as follows, where λ smo represents the penalty coefficient matrix,

[0095]

[0096] Among them J end Represents the terminal pose constraint. If the final target point is not reached, the penalty coefficient λ of this optimization objective end It can be set appropriately smaller, as follows

[0097] J end =λ end ||z i -z e || (5b)

[0098] The mobile robot also needs to satisfy the arc constraint as much as possible, but this constraint is not a mandatory constraint and is a very loose constraint, as shown below, where λ kin is the constraint coefficient,

[0099]

[0100] The mobile robot also needs to satisfy the no-collision constraint as much as possible and try to drive in the middle of adjacent obstacles. Therefore, the constraints are set as follows, where λ v represents the constraint coefficient, R i represents the posture matrix of the mobile robot, o j Indicates obstacles, p i Represents the mobile robot's position information.

[0101]

[0102] The meanings of other characters in the above variables are as follows: i represents the pose vector, p represents the position information, is a two-dimensional real number space, representing a point on the plane, S 1 represents the parameterization of the circle, which is often used to describe rotation or direction. h(*,*) represents the kinematic constraint function of the mobile robot.

[0103]

[0104] p = [x i ,y i ] T (6a)

[0105] Δz i =z i+1 -z i (6b)

[0106] δz i,i+1 =Δz i+1 -Δz i (6c)

[0107]

[0108] h(z i ,z i+1 )=|(V θ,i +V θ,i+1 )×V p,i |2 (6e)

[0109] V θ,i =[cos(θ i ), sin(θ i )] T (6f)

[0110] V p,i =[x i+1 -x i ,y i+1 -y i ] T (6g)

[0111] In the above steps, the planning of the pose point is completed, which is a discrete path point. Based on the optimized pose sequence, the cubic polynomial fitting and interpolation based on energy optimization are used. The formulas of the cubic polynomial to be fitted in the XY direction are respectively recorded as

[0112]

[0113] Where s(τ), They are respectively expressed as follows:

[0114] d a =[a0,a1,a2,a3],d b =[b0,b1,b2,b3] (7a)

[0115] s(τ)=[1,τ,τ 2 ,τ 3 ],τ∈[0,1] (7b) The starting and ending points of the trajectory lines of adjacent postures need to meet the requirements of position and posture coincidence. Therefore, there are the following constraints:

[0116]

[0117] The above η k , γ k is the relaxation factor, then the optimization variable can be written as follows,

[0118]

[0119] Then (7c) and (7d) can be written as follows

[0120] AX k =0 (7f)

[0121] Then there is

[0122]

[0123] Use energy optimization as the objective function as follows

[0124]

[0125] Similarly, there are

[0126]

[0127] sxt k =0

[0128] Where G can be expressed as follows

[0129]

[0130] The above-mentioned target equation is a quadratic problem, namely a QP problem, and thus can be solved quickly, with a computation time of less than 2 ms. In this step, the polynomial fitting process of adjacent pose points is completed.

[0131] Specific implementation of step S600

[0132] If the energy optimal polynomial curve fitted in step S500 collides with an obstacle, the pose sampling density at the collision point is increased, and steps S400 and S500 are performed again. If no collision occurs, the current planning ends, and the coefficient parameters of the polynomial curve and the optimized pose sequence are used as the path planning results of the mobile robot and are transmitted to the motion controller of the mobile robot.

[0133] Figures 9 to 12 The robot trajectory planning method using the two-step optimization strategy of the present invention is verified by using simulation scenarios and actual scenarios. Fig. 9 It is a planning diagram in a random obstacle environment; Fig.10 This is a schematic diagram of planning in a channel environment with a width of 0.7 m according to an example of the present invention; Fig.11 is a schematic diagram of planning in a narrow curved passage according to an example of the present invention; Fig.12 is a diagram of a physical test scene of a mobile robot in an example of the present invention, which verifies the safe real-time navigation of the mobile robot in a narrow space.

[0134] It should be appreciated that the method steps in the embodiments of the present invention can be implemented or implemented by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer readable memory. The method can use standard programming techniques. Each program can be implemented in a high-level process or object-oriented programming language to communicate with a computer system. However, if necessary, the program can be implemented in an assembly or machine language. In any case, the language can be a compiled or interpreted language. In addition, the program can be run on a programmed ASIC for this purpose.

[0135] Furthermore, the operations of the processes described herein may be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The processes described herein (or variations and / or combinations thereof) may be performed under the control of one or more computer systems configured with executable instructions, and may be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) that is executed collectively on one or more processors, by hardware, or a combination thereof. The computer program includes a plurality of instructions that may be executed by one or more processors.

[0136] Further, the method can be implemented in any type of computing platform that is operably connected to a suitable computer, including but not limited to a personal computer, a minicomputer, a mainframe, a workstation, a network or distributed computing environment, a separate or integrated computer platform, or in communication with a charged particle tool or other imaging device, etc. Various aspects of the present invention can be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into a computing platform, such as a hard disk, an optical read and / or write storage medium, a RAM, a ROM, etc., so that it can be read by a programmable computer, and when the storage medium or device is read by the computer, it can be used to configure and operate the computer to perform the process described herein. In addition, the machine-readable code, or portions thereof, can be transmitted via a wired or wireless network. When such media includes instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor, the invention described herein includes these and other different types of non-transitory computer-readable storage media. When programmed according to the methods and techniques of the present invention, the present invention can also include the computer itself.

[0137] The computer program can be applied to input data to perform the functions described herein, thereby converting the input data to generate output data stored in a non-volatile memory. The output information can also be applied to one or more output devices such as a display. In a preferred embodiment of the present invention, the converted data represents physical and tangible objects, including specific visual depictions of physical and tangible objects produced on the display.

[0138] It should be noted that, if not otherwise specified, when a certain feature is referred to as "fixed", "connected" to another feature, it can be directly fixed, connected to another feature, or indirectly fixed, connected to another feature. It should be understood that, although the terms first, second, third, etc. may be used to describe various elements herein, these elements should not be limited to these terms. These terms are only used to distinguish the same type of elements from each other. For example, without departing from the scope of this article, the first element may also be referred to as the second element, and similarly, the second element may also be referred to as the first element. The use of any and all examples or exemplary language ("for example", "such as", etc.) provided herein is intended only to better illustrate embodiments of the present invention, and unless otherwise required, will not impose limitations on the scope of the present invention.

[0139] The above is only a preferred embodiment of the present invention. The present invention is not limited to the above implementation. As long as the technical effect of the present invention is achieved by the same means, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the scope of protection of the present invention. Within the scope of protection of the present invention, its technical scheme and / or implementation method may have various modifications and changes.

Claims

1. A path planning method for a mobile robot, characterized in that: The method comprises the following steps: S100, acquiring geometric data of a maximum inscribed circle according to the chassis shape of the mobile robot, and configuring a set of control points of a preset number at an initial sampling resolution to be associated with a trajectory line of path planning of the mobile robot; S200, updating the collision data of the mobile robot, constructing a collision avoidance constraint, and optimizing the positions of a set of control points based on a first objective function of minimum path length consumption under a collision-free condition, and fitting the control points obtained by the optimization calculation into an updated trajectory line; S300, determining that the maximum inscribed circle of the mobile robot does not collide with obstacles when moving along all strokes of the updated trajectory line, then obtaining the polynomial curve trajectory geometry data under the inscribed circle constraint model as the initial value of the refined optimized trajectory, otherwise, adding more control points in the collision area, and returning to step S200; S400, establishing a Euclidean distance field model of the mobile robot according to the precise geometric shape data obtained from the chassis shape of the mobile robot, adding a collision risk area, and constructing a collision constraint; S500, using the curve generated by the polynomial curve trajectory geometry data as the connecting line of the path points of the mobile robot, establishing the posture sequence of the mobile robot in the path points as the optimization variable, and solving the optimized posture sequence based on the optimization equation of the second objective function of minimum energy consumption, and fitting the posture sequence obtained by the optimization solution into an updated polynomial curve; S600, determine that the precise geometric shape of the mobile robot does not collide with obstacles along all the strokes of the fitted polynomial curve, then obtain the coefficient parameters of the polynomial curve and the optimized posture sequence as the path planning result of the mobile robot; otherwise, increase the posture sampling density in the collision area to generate a posture sequence, and return to step S400.

2. The method according to claim 1, wherein: The step S100 includes: Set the risk area and provide the risk penalty value calculation function f through the following formula obs , used for the optimization calculation of the first objective function of minimum energy consumption, Where, ρ represents the distance between the center of the chassis of the mobile robot and the edge of the obstacle, r m represents the maximum inscribed circle radius of the chassis shape of the mobile robot, c represents the maximum value of the risk area, δ represents the maximum distance from the edge of the risk area to the edge where the collision occurs, If <r m Then the mobile robot is in the collision area if r m < <r m +δ then the mobile robot is in the risk area.

3. The method according to claim 1 or 2, wherein: The step S100 further includes: Preliminary curve sampling is performed along the paths of the mobile robot's movement starting point, movement target point and obstacle detour points to evenly configure a group of control points, the initial configuration number of which is substantially equal to the total number of points of obstacle detour, movement starting point and movement target point.

4. The method according to claim 1 or 2, wherein: The step S200 includes the following steps: S210, the first objective function with minimum energy consumption is obtained by the following formula The gradient descent algorithm is used to optimally solve the position vector q of the control point. Among them, J b represents the collision constraint, J sm Indicates the drastic change of the control point, J le represents the length constraint of the trajectory line, λ b , sm , le They represent the constraint coefficient, smoothness coefficient and path length coefficient of collision avoidance, respectively. obs represents the risk penalty value calculation function, o j Represents the obstacle position vector. i, j, n, and m are natural numbers and are used for subscript and subscript counting of parameters in the formula.

5. The method according to claim 4, wherein: The step S200 further includes the following steps: S220, fitting the control points obtained by optimization calculation into a spline curve p by the following formula i (τ) and as the trajectory, Among them, τ is the function variable used to generate the spline curve, τ∈[0,1],n p is the polynomial order, The constant matrix representing the spline curve, s(τ) is the vector of polynomial terms that constitute the spline curve, π i is the control point matrix, i is a natural number used to count the subscripts and subscripts of the parameters in the formula, i≤n q -n p , n q is the number of control points.

6. The method according to claim 1, wherein: The step S400 includes: Based on the Euclidean distance field model, the collision risk value calculation function is provided by the following formula To establish collision constraints, solve the optimization equation of the second objective function for minimum energy consumption, Among them, o represents the obstacle vector, c represents the maximum value of the risk area, represents the Euclidean distance field model operation, represents the shape space of the mobile robot chassis, represents the expanded shape space of the mobile robot chassis after superimposing the risk area, Indicates that the obstacle is within the collision range. Indicates that the obstacle is within the risk area, δ indicates the maximum distance from the edge of the risk area to the edge where the collision occurs, and min{||op||} indicates the shortest distance from the obstacle to the boundary.

7. The method according to claim 1 or 6, wherein: The step S500 includes the following steps: S510, the second objective function with minimum energy consumption is calculated by the following formula Optimally solve the posture sequence z of the mobile robot, J end =λ end ||from i -With e || Among them, J smo represents the smoothness constraint, J end represents the terminal pose constraint, J kin represents the kinematic constraint, J v represents no collision constraint, λ smo represents the smoothness penalty coefficient matrix, λ end represents the penalty coefficient of the optimization objective, λ kin represents the arc constraint coefficient, λ v represents the non-collision constraint coefficient, δ represents the maximum distance from the edge of the risk area to the edge where the collision occurs, h represents the kinematic constraint function of the mobile robot, and z i represents the pose vector, z e represents the end pose vector, represents the collision risk value calculation function, R i represents the posture matrix of the mobile robot, o j represents the position vector of the obstacle, p i Represents the position vector of the mobile robot. i and j are natural numbers and are used for subscript and subscript counting of parameters in the formula.

8. The method according to claim 7, wherein: The step S500 The following steps are involved: S520, performing cubic polynomial curve fitting in the X and Y directions on the posture sequence z obtained by optimization using the following formula: Where τ is the function variable used to generate the cubic polynomial curve, s(τ) is the vector of polynomial terms that constitute the cubic polynomial curve, and d a and d b are the parameter matrices of the polynomial terms in the X and Y directions respectively; Among them, the constraint condition AX is given according to the requirement that the starting and ending points of the trajectory line of adjacent posture points in the posture sequence of the mobile robot need to meet the requirement that the position and posture coincide. k = 0, and the energy optimization is used as the objective function through the following formula Optimally solve the optimization variable X of the cubic polynomial curve k , s.t.AX k =0 in, η k and γ k is the relaxation factor coefficient, B and G are constant matrices, and k is the formula letter subscript.

9. A computing device comprising a memory and a processor, characterized in that: The method according to any one of claims 1 to 8 is implemented when the processor executes the computer application stored in the memory.

10. A mobile robot, characterized in that: include: Environmental sensors, used to obtain the location and shape data of obstacles; Posture sensor, used to feedback the position and posture data of the mobile robot; According to the computing device according to claim 9, the computing device is connected to the environmental sensor and the posture sensor respectively.

Images