CN106681135B - Cable wiring path searching method based on mixed water drop algorithm - Google Patents

Abstract

The invention discloses a cable path automatic planning method based on an improved mixed water drop algorithm, which mainly solves the problems of low algorithm efficiency, unsmooth path and the like in the prior art. The planning steps are as follows: modeling a cable wiring environment; optimizing parameters in an intelligent water droplet algorithm by utilizing a particle swarm algorithm, and providing guidance for movement of water droplets; improving the moving mode of the water drops in the water drop algorithm and optimizing the path; outputting a cable path according to the optimization result; the method comprehensively considers the stability and robustness of the algorithm, improves the searching efficiency of the algorithm, avoids the interference phenomenon between the cable and the barrier, and reduces the length of the cable required by wiring.

Description

Cable wiring path searching method based on mixed water drop algorithm

Technical Field

The invention relates to the field of automatic wiring, in particular to a cable wiring path searching method based on a mixed water drop algorithm.

Background

The searching of the cable wiring path is one of key technologies for realizing automatic wiring of cables, and the cable wiring path which is convenient, collision-free, smooth and attached to the wall surface can be found out quickly, so that the automatic wiring efficiency of the cables is improved, and the reliability of electromechanical products is also improved.

At present, commonly used path searching methods include a visual graph method, a heuristic graph searching algorithm, an artificial potential field method and the like, the algorithms have respective advantages and disadvantages, for example, the artificial potential field method has good real-time performance, but has the defects that a trap area exists, a path cannot be found between similar obstacles and the like, and the algorithm is more suitable for solving the single-target optimization problem. In recent decades, as the research of artificial intelligence algorithms has been advanced, many intelligent algorithms are also used in path planning of mobile robots, including fuzzy logic and reinforcement learning algorithms, neural networks, genetic algorithms, ant colony algorithms, and the like. These algorithms have their own advantages, but have many problems, such as too slow convergence speed, easy falling into local optimum, too large search space, etc.

Although many algorithms are improved at the present stage, a better cable wiring path can be found, the problems of more iteration times, overlong operation time and the like still exist, the problems cannot meet the requirement of automatic cable wiring, and in the obtained path, the cable path is in a suspended state, so that the stability of an electromechanical product is seriously influenced, and meanwhile, the working reliability of the electromechanical product is reduced.

Disclosure of Invention

Aiming at the defects of the prior art, the invention provides a cable wiring path searching algorithm based on a mixed water drop algorithm, the method has the advantages of less iteration times and short operation time, can meet the requirement of automatic wiring of cables, does not influence the stability of an electromechanical product, and improves the working reliability of the electromechanical product.

The technical scheme for realizing the purpose of the invention is as follows:

a cable wiring path searching method based on a mixed water drop algorithm comprises the following steps:

1) in the wiring space, selecting a wall surface for laying the cable, and simplifying the wiring space into a rectangular wiring plane in a plane expansion mode;

2) carrying out grid map modeling on a rectangular wiring plane;

3) taking a particle swarm algorithm as a coarse searching stage, carrying out rapid path optimization, and updating soil distribution through a plurality of groups of obtained suboptimal paths;

4) an improved intelligent water drop algorithm is adopted as a fine search stage to carry out path optimization;

5) taking the nodes and the bending points in the path as control points, and converting two-dimensional control point coordinates into three-dimensional control point coordinates in the wiring space according to the plane unfolding mode in the step 1);

6) optimizing a path by adopting a third-order B spline curve;

through the steps, parameters in the intelligent water droplet algorithm are optimized by utilizing the particle swarm algorithm, and guidance is provided for the movement of water droplets; improving the moving mode of the water drops in the water drop algorithm and optimizing the path; finally, a B-spline curve is adopted to optimize a path, the stability and robustness of the algorithm are comprehensively considered, and a wiring path which is convenient, collision-free, smooth and attached to a wall surface can be quickly found out. The cabling is completed by laying the cables on the path.

In the step 1), in the wiring space, selecting a wall surface for laying the cable, and simplifying the wiring space into a rectangular wiring plane in a plane expansion mode:

firstly, if a plane is a non-rectangular plane simplified after plane expansion, the plane is expanded into a rectangular plane;

secondly, the extension part is regarded as an obstacle;

finally, the plane is divided into two regions: barrier regions and non-barrier regions.

In step 2), performing grid map modeling on the rectangular wiring plane:

firstly, the grid map consists of obstacle grids and non-obstacle grids;

secondly, any grid containing obstacles is regarded as an obstacle grid;

and finally, if the starting point and the ending point of the wiring are positioned in the same row in the grid map, interchanging the rows and the columns of the grid map.

In step 3), the particle swarm algorithm is used as a coarse search stage and rapid path optimization is carried out, and soil distribution is updated through multiple sets of obtained suboptimal paths, and the method specifically comprises the following steps:

first, initial parameters (population size N, inertial weight w, learning factor c) are set₁And c₂Maximum number of iterations Ic, maximum speed V_maxAnd a minimum value of velocity V_min)；

Then, a position vector x of the particle is randomly generated_i＝(x_i1,x_i2,…,x_iD) And velocity vector v_i＝(v_i1,v_i2,…,v_iD) And initializing a locally optimal solution p of the particle_pAnd a global optimal solution p_g；

Entering iteration and updating the position information and the speed information of all particles according to the following two formulas:

wherein i is 1,2, …, N, D is 1,2, …, D, w is inertial weight, c1 and c2 are learning factors, r1 and r2 are [0,1]Are relatively independent and uniformly distributed random numbers p_p＝(p_p1,p_p2,…,p_pD) For the local optimal solution of the particle, i.e. the optimal solution of the particle in the last iteration, p_g＝(p_g1,p_g2,…,p_gD) Is the global optimal solution of the particle, i.e., the optimal solution of the particle before this iteration.

At the same time, the fitness f (x) of each particle is calculated_i) If f (x)_i) Is the minimum value in this iteration, then p_p＝x_iIf f (x) is present_i)<f(p_g) Then p is_g＝x_iWherein the fitness function f is:

in the formula (x)_j,y_j) Is the coordinate of the path node j, n is the number of path points, num_jAnd pun is a penalty factor for the number of obstacle grids touched by the path between the node j and the node j-1, so that the path touching the obstacle is endowed with a poor evaluation value.

Finally, the next iteration is enteredThe local optimal solution p is updated according to the formulas (1), (2) and (3)_pAnd global maximum solution p_gAnd until the iteration times reach the maximum or meet the precision requirement.

In the step 4), an improved intelligent water drop algorithm is adopted as a fine searching stage to carry out path optimization,

firstly, setting the number W of water drops, the maximum iteration number Is and the initial soil amount Its, initializing the soil at each position in the grid map, and updating the soil by utilizing T groups of suboptimal solutions obtained by a PSO algorithm, wherein the T groups of suboptimal solutions are as follows:

in the formula of oil_iThe amount of soil at a certain node i; pos_iIs the location of the node; path_subThe position of each node in the path in the D group of paths; eup is the shortest distance between the central points of two adjacent grids; lambda [ alpha ]_iIndicates λ in the T-set sub-optimal solution_iThe block solution contains node i, and the parameters δ, α and β are positive numbers.

Then, all the water droplets were placed at the wiring start point.

Using roulette method to select path point and recording path node information, the formula is as follows:

in the formula of oil_iThe amount of soil at a certain node; pos_iIs the location of the node; eup is the shortest distance between the central points of two adjacent grids; p is the probability of selecting the next node; n is the set of next optional grids; function dis (pos)_i,pos_j) Is composed ofEuclidean metrics between node i and node j; ρ is a positive number less than 1; a is_f,，b_fAnd c_fAre all positive numbers.

When the water drop moves to another node j, the speed of the water drop is updated:

middle vel of the formula_jIs the speed at which the drop moves to node j; soil_jThe amount of soil at node j; a is_vIs a positive number.

Meanwhile, the soil of the moved node and the soil carried by the water drop at the moment are updated, and the following two formulas are adopted:

middle pos_iIs the position of node i; function dis (pos)_i,pos_j) Is the Euclidean metric between node i and node j; soil_jThe amount of soil at node j; rho₁And a_sAre all positive numbers.

Before all water drops move to the wiring end point, the formulas (6), (7), (8), (9) and (10) are repeated, the fitness of the formed path of each water drop is calculated, and the local optimal solution p is updated_pAnd a global optimal solution p_g。

After all water drops moved to the wiring end point, the global soil volume was updated:

in the formula of oil_iThe amount of soil at node i; pos_iIs the position of node i;

the amount of soil carried by the water droplets to form a locally optimal solution; p is a radical of_pIs a locally optimal solution, and p_gIs a global optimal solution; num number of grids in the grid map; local cognitive factor ρ₂And global cognition factor ρ₃Are all positive numbers, and ρ₃≥ρ₂。

After the total local soil quantity is updated according to the formula (11), the algorithm completes one iteration. Thereafter the algorithm will go to the next iteration and all drops will reset at the start of the wiring and move again in the above described way (equation (6) to equation (11)) until the number of iterations reaches the maximum or the accuracy requirement is met.

In the step 5), taking the nodes and the bending points in the path as control points, and converting two-dimensional control point coordinates into three-dimensional control point coordinates in the wiring space according to the plane unfolding mode in the step 1);

in step 6), a third-order B-spline is adopted to optimize a path:

obtaining the control point p in the step 5) first₀(x₀,y₀,z₀),...,p_n(x_n,y_n,z_n) Determining the order k of the B spline curve to be 3, and defining the support interval T of the spline curve to be T₀,t₁,...,t_n+k]The following formula:

the basis function N of the spline curve is then calculated_i,k(t) the following two formulae:

wherein t is a variable in a spline curve equation; t ═ T₀,t₁,...,t_n+k]Is the support interval of the spline curve.

Finally, the equation of the third-order B spline curve can be obtained as follows:

in the formula, t is a variable of a curve equation; n is a radical of_i,k(t) is a basis function in the B-spline curve;

k is the curve order of 3.

Has the advantages that:

the invention provides a cable wiring path searching method based on a mixed water drop algorithm, which comprises the following steps:

(1) the wiring space is simplified into a wiring plane by selecting a cable laying wall surface, and after a cable path is solved, two-dimensional coordinates of a cable control point on the plane are converted into three-dimensional coordinates of the wiring space, so that the wiring path is attached to the wall surface, and the condition that the cable is suspended is avoided;

(2) the grid map modeling mode is adopted, so that the method has the advantages of high precision, easiness in implementation and the like;

(3) and the particle swarm algorithm and the improved intelligent water drop algorithm are fused. The particle swarm algorithm has the advantages that the required time is short compared with that of an intelligent water drop algorithm, so that not only is too much time not consumed before the intelligent water drop algorithm, but also the whole running speed is not influenced, and pheromone feedback of initial soil distribution can be provided for the intelligent water drop algorithm, so that the running time of the intelligent water drop algorithm can be reduced, and the effects of reducing the searching time and obtaining the optimized path can be achieved;

(4) the Euclidean metric is used as a factor influencing the movement of the water drop selection node, so that the water drop moves towards a better-positioned area as far as possible, and the convergence rate of the algorithm is increased;

(5) a third-order B-spline curve is adopted, so that the cable path is smoother, and a large number of bending points with large bending angles in the cable path are avoided.

Drawings

FIG. 1 is a flow chart of grid map creation;

FIG. 2 is a flow chart of cable routing path search based on the mixed water droplet algorithm;

FIG. 3 is an automatic cabling of cables in a cabling space accomplished by the method of the present invention;

fig. 4 simplifies the routing space flow diagram.

Detailed Description

The invention will be further elucidated with reference to the drawings and examples, without however being limited thereto.

Examples

A cable wiring path searching method based on a mixed water drop algorithm comprises the following steps:

1) in the wiring space, selecting a wall surface for laying the cable, and simplifying the wiring space into a rectangular wiring plane in a plane expansion mode;

2) carrying out grid map modeling on a rectangular wiring plane;

3) taking a particle swarm algorithm as a coarse searching stage, carrying out rapid path optimization, and updating soil distribution through a plurality of groups of obtained suboptimal paths;

4) an improved intelligent water drop algorithm is adopted as a fine search stage to carry out path optimization;

5) taking the nodes and the bending points in the path as control points, and converting two-dimensional control point coordinates into three-dimensional control point coordinates in the wiring space according to the plane unfolding mode in the step 1);

6) optimizing a path by adopting a third-order B spline curve;

through the steps, parameters in the intelligent water droplet algorithm are optimized by utilizing the particle swarm algorithm, and guidance is provided for the movement of water droplets; improving the moving mode of the water drops in the water drop algorithm and optimizing the path; finally, a B-spline curve is adopted to optimize a path, the stability and robustness of the algorithm are comprehensively considered, and a wiring path which is convenient, collision-free, smooth and attached to a wall surface can be quickly found out. The cabling is completed by laying the cables on the path.

In the step 1), in the wiring space, selecting a wall surface for laying the cable, and simplifying the wiring space into a rectangular wiring plane in a plane expansion mode:

as shown in fig. 1, in a simplified manner of wiring space. Wherein P is₁As a starting point of wiring, P₂The key points of wiring are; s_o，S₁And S₂All are wall surfaces for laying cables; o is₁Is an obstacle, and O₂Is an extended area and is considered an obstacle; finally, the black area represents an obstacle, and the white area represents a non-obstacle area.

Firstly, if a plane is a non-rectangular plane simplified after plane expansion, the plane is expanded into a rectangular plane;

secondly, the extension part is regarded as an obstacle;

finally, the plane is divided into two regions: barrier regions and non-barrier regions.

In step 2), performing grid map modeling on the rectangular wiring plane:

as shown in fig. 2, a grid modeling approach is used. Wherein the circular ring represents a wiring starting point, and the rectangular ring represents a wiring end point; the middle black grid is a barrier grid; the white grid is a non-obstacle grid; the distance between any two grids is the line connecting their center points.

Firstly, the grid map consists of obstacle grids and non-obstacle grids;

secondly, any grid containing obstacles is regarded as an obstacle grid;

and finally, if the starting point and the ending point of the wiring are positioned in the same row in the grid map, interchanging the rows and the columns of the grid map.

In step 3), the particle swarm algorithm is used as a coarse search stage and rapid path optimization is carried out, and soil distribution is updated through multiple sets of obtained suboptimal paths, and the method specifically comprises the following steps:

first, initial parameters (population size N, inertial weight w, learning factor c) are set₁And c₂Maximum number of iterations Ic, maximum speed V_maxAnd a minimum value of velocity V_min)；

Then, a position vector x of the particle is randomly generated_i＝(x_i1,x_i2,…,x_iD) And velocity vector v_i＝(v_i1,v_i2,…,v_iD) And initializing a locally optimal solution p of the particle_pAnd a global optimal solution p_g；

Entering iteration and updating the position information and the speed information of all particles according to the following two formulas:

wherein i is 1,2, …, N, D is 1,2, …, D, w is inertial weight, c1 and c2 are learning factors, r1 and r2 are [0,1]Are relatively independent and uniformly distributed random numbers p_p＝(p_p1,p_p2,…,p_pD) For the local optimal solution of the particle, i.e. the optimal solution of the particle in the last iteration, p_g＝(p_g1,p_g2,…,p_gD) Is the global optimal solution of the particle, i.e., the optimal solution of the particle before this iteration.

At the same time, the fitness f (x) of each particle is calculated_i) If f (x)_i) Is the minimum value in this iteration, then p_p＝x_iIf f (x) is present_i)<f(p_g) Then p is_g＝x_iWherein the fitness function f is:

in the formula (x)_j,y_j) Is the coordinate of the path node j, n is the number of path points, num_jAnd pun is a penalty factor for the number of obstacle grids touched by the path between the node j and the node j-1, so that the path touching the obstacle is endowed with a poor evaluation value.

Finally, entering next iteration and repeating the formulas (1), (2) and (3), and updating the local optimal solution p_pAnd global maximum solution p_gAnd until the iteration times reach the maximum or meet the precision requirement.

In the step 4), an improved intelligent water drop algorithm is adopted as a fine searching stage to carry out path optimization,

firstly, setting the number W of water drops, the maximum iteration number Is and the initial soil amount Its, initializing the soil at each position in the grid map, and updating the soil by utilizing T groups of suboptimal solutions obtained by a PSO algorithm, wherein the T groups of suboptimal solutions are as follows:

in the formula of oil_iThe amount of soil at a certain node i; pos_iIs the location of the node; path_subThe position of each node in the path in the D group of paths; eup is the shortest distance between the central points of two adjacent grids; lambda [ alpha ]_iIndicates λ in the T-set sub-optimal solution_iThe block solution contains node i, and the parameters δ, α and β are positive numbers.

Then, all the water droplets were placed at the wiring start point.

Using roulette method to select path point and recording path node information, the formula is as follows:

in the formula of oil_iThe amount of soil at a certain node; pos_iIs the location of the node; eup is the shortest distance between the central points of two adjacent grids; p is the probability of selecting the next node; n is the set of next optional grids; function dis (pos)_i,pos_j) Is the Euclidean metric between node i and node j; ρ is a positive number less than 1; a is_f,，b_fAnd c_fAre all positive numbers.

When the water drop moves to another node j, the speed of the water drop is updated:

middle vel of the formula_jIs the speed at which the drop moves to node j; soil_jThe amount of soil at node j; a is_vIs a positive number.

Meanwhile, the soil of the moved node and the soil carried by the water drop at the moment are updated, and the following two formulas are adopted:

middle pos_iIs the position of node i; function dis (pos)_i,pos_j) Is the Euclidean metric between node i and node j; soil_jThe amount of soil at node j; rho₁And a_sAre all positive numbers.

Before all water drops move to the wiring end point, the formulas (6), (7), (8), (9) and (10) are repeated, the fitness of the formed path of each water drop is calculated, and the local optimal solution p is updated_pAnd a global optimal solution p_g。

After all water drops moved to the wiring end point, the global soil volume was updated:

in the formula of oil_iThe amount of soil at node i; pos_iIs the position of node i;

the amount of soil carried by the water droplets to form a locally optimal solution; p is a radical of_pIs a locally optimal solution, and p_gIs a global optimal solution; num number of grids in the grid map; local cognitive factor ρ₂And all ofLocal cognition factor rho₃Are all positive numbers, and ρ₃≥ρ₂。

After the total local soil quantity is updated according to the formula (11), the algorithm completes one iteration. Thereafter, the algorithm will proceed to the next iteration, and all the water drops will be reset at the starting point of the wiring and will move again in the above equations (6) to (11) until the number of iterations reaches the maximum or the accuracy requirement is met.

In the step 5), taking the nodes and the bending points in the path as control points, and converting two-dimensional control point coordinates into three-dimensional control point coordinates in the wiring space according to the plane unfolding mode in the step 1);

in step 6), a third-order B-spline is adopted to optimize a path:

as shown in fig. 3, which is an example of the wiring path optimized by using the third-order B-spline. The wiring space is a box body of 10dm multiplied by 18dm multiplied by 20dm, the barrier is an object surrounded by a dotted line, the short cylinder surrounded by a solid line represents a connector, the starting and ending ends of the cable connecting wiring can be obviously seen, the barrier is avoided, and the cable connecting wiring is pasted on the wall surface;

obtaining the control point p in the step 5) first₀(x₀,y₀,z₀),...,p_n(x_n,y_n,z_n) Determining the order k of the B spline curve to be 3, and defining the support interval T of the spline curve to be T₀,t₁,...,t_n+k]The following formula:

the basis function N of the spline curve is then calculated_i,k(t) the following two formulae:

wherein t is a variable in a spline curve equation; t ═ T₀,t₁,...,t_n+k]Is the support interval of the spline curve.

Finally, the equation of the third-order B spline curve can be obtained as follows:

in the formula, t is a variable of a curve equation; n is a radical of_i,k(t) is a basis function in the B-spline curve;

k is the curve order of 3.

As shown in fig. 4: the flow of the mixed water drop algorithm for solving the wiring path is as follows:

s101 begins

S102 initialization parameters

S103, solving T-group suboptimal path by adopting particle swarm optimization

S104, updating the global soil quantity according to the T-group suboptimal path

S105 iteration number increase

S106 selecting the water drops which are not moved yet

S107 water drop selection node movement

S108, updating the soil at the node, the moving speed of water drops and the soil carried by the water drops

S109 judges whether or not the water droplet moves to the wiring end point

S110, judging whether all water drops move to the wiring end point

S111 updating local optimal solution and global optimal solution

S112 updating the global soil

S113 judges whether or not an iteration end condition is satisfied

S114 obtaining control point coordinates

S115 adopts a third-order B spline curve to optimize a path

S116 output cable routing path

And S117 ends.

Claims (4)

1. A cable wiring path searching method based on a mixed water drop algorithm is characterized by comprising the following steps:

1) in the wiring space, selecting a wall surface for laying the cable, and simplifying the wiring space into a rectangular wiring plane in a plane expansion mode;

2) carrying out grid map modeling on a rectangular wiring plane;

3) taking a particle swarm algorithm as a coarse searching stage, carrying out rapid path optimization, and updating soil distribution through a plurality of groups of obtained suboptimal paths;

4) the method comprises the steps of adopting a mixed water drop algorithm as a fine search stage to conduct path optimization, firstly, setting the quantity W of water drops, the maximum iteration times Is and the initial soil quantity Its, initializing soil at each position in a grid map, and utilizing T groups of suboptimal solutions obtained by a PSO algorithm to update the soil, wherein the T groups of suboptimal solutions are as follows:

in the formula: its is the initial soil mass; pos_iIs position i; sol_iThe displacement position is the soil amount at the position i; lambda [ alpha ]_iPositive number for controlling soil change, T is number of fine searches, delta, α, β are real numbers for regulating initial soil quantity, eup is shortest distance between adjacent positions, path_subA path set for fine search; initsoil is Its initial soil amount;

then, all the water drops are placed at the wiring starting point,

using roulette method to select path point and recording path node information, the formula is as follows:

in the formula: soil_ijIs the amount of soil at coordinate (i, j); pos_i、pos_jPositions i and j; dis (pos)_i,pos_j) Is the Euclidean distance between location i and location j; rho is a coefficient; a is_fPositive, affecting the probability of a drop moving from location i to the next location j; b_fIs a positive number;

when the water drop moves to another node j, the speed of the water drop is updated:

in the formula: a is_vFor controlling the rate of change of the speed of the water droplets; vel_i,vel_jThe speeds of water drops at the i and j nodes respectively; soil_jThe amount of soil at point j;

at the same time, the soil of the moved node is updated, and the soil carried by the water drop at this time:

in the formula: a is_sFor controlling the rate of change of the speed of the water droplets; vel_jIs the velocity of the water droplet at the j node; pos_i、pos_jPositions i and j; soil_jThe amount of soil at point j; dis (pos)_i,pos_j) Is the Euclidean distance, ρ, between location i and location j₁Is a coefficient;

in the formula: soil^IWDIs the amount of soil, pos, carried by the water droplets_i、pos_jPositions i and j; vel_jThe velocity at point j; dis (pos)_i,pos_j) Is the Euclidean distance between location i and location j; a is_sIs a positive number for controlling the rate of change of the soil;

before all water droplets move to the wiring end point, equations (6), (7), (8), (9) and (10) are repeated, and each is calculatedThe fitness of the path formed by the water drops is updated, and the local optimal solution p is updated_pAnd global maximum solution p_g；

After all water drops moved to the wiring end point, the global soil volume was updated:

in the formula: sol_iRepresents the amount of soil at location i; p is a radical of_pRepresents a locally optimal solution; p is a radical of_gRepresents the global maximum solution;

the optimal value of the soil quantity carried by water drops is obtained; rho₂，ρ₃，ρ₄Is a coefficient; num is the number of water droplets; after the total local soil quantity is updated according to the formula (11), the algorithm finishes one iteration, then the algorithm enters the next iteration, all water drops are reset at the wiring starting point, and the movement is carried out again according to the formula (6) to the formula (11) until the iteration frequency reaches the maximum or the precision requirement is met;

5) taking the nodes and the bending points in the path as control points, and converting two-dimensional control point coordinates into three-dimensional control point coordinates in the wiring space according to the plane unfolding mode in the step 1);

6) optimizing a path by adopting a third-order B spline curve; acquiring the control point p in the step 5) firstly₀(x₀,y₀,z₀)…p_n(x_n,y_n,z_n) Determining the order k of the B spline curve to be 3, and defining the support interval T of the spline curve to be T₀,t₁,…,t_n+k]The following formula:

in the formula: i represents the curve control point i, N represents the total number of control points, k is the order of the spline and then the basis function N of the spline is calculated_i,k(t) the following two formulae:

in the formula: basis function N of control point i_i,1,t_iAnd t_i+1Respectively support zones i and i +1,

in the formula: t is t_i、t_i+1、t_i+k-1、t_i+kRespectively supporting intervals i, i +1, i + k, i + k-1; n is a radical of_i,k-1(t) and N_i+1,k-1(t) is the basis function of point i and point i + 1;

finally, the equation of the third-order B spline curve can be obtained as follows:

in the formula: i represents a curve control point i; basis function N of control point i_i,k(t)；P_iIs the position vector of control point i.

2. The method for searching a cable wiring path based on the mixed water drop algorithm according to claim 1, wherein in the step 1), the wall surface for cable laying is selected in the wiring space, and the wiring space is simplified into a rectangular wiring plane in a planar spreading manner:

firstly, if a plane is a non-rectangular plane simplified after plane expansion, the plane is expanded into a rectangular plane;

secondly, the extension part is regarded as an obstacle;

finally, the plane is divided into two regions: barrier regions and non-barrier regions.

3. The cable wiring path searching method based on the mixed water drop algorithm according to claim 1, wherein in the step 2), the grid map modeling is performed on the rectangular wiring plane:

firstly, the grid map consists of obstacle grids and non-obstacle grids;

secondly, any grid containing obstacles is regarded as an obstacle grid;

and finally, if the starting point and the ending point of the wiring are positioned in the same row in the grid map, interchanging the rows and the columns of the grid map.

4. The cable wiring path searching method based on the mixed water drop algorithm according to claim 1, wherein in the step 3), the particle swarm algorithm is used as a coarse searching stage to perform fast path optimization, and the soil distribution is updated through the obtained multiple sets of suboptimal paths, which specifically comprises:

firstly, the population size N, the inertia weight w and the learning factor c are calculated₁And c₂Maximum number of iterations I_CMaximum value of velocity V_maxAnd a minimum value of velocity V_minSetting as initial parameters; then, a position vector x of the particle is randomly generated_i＝(x_i1,x_i2,…,x_iD) And velocity vector v_i＝(v_i1,v_i2,…,v_iD) And initializing a locally optimal solution p of the particle_pAnd a global optimal solution p_g；

Entering iteration and updating the position information and the speed information of all particles according to the following two formulas:

at the same time, the adaptation f (x) of each particle is calculated_i) If f (x)_i) Is the minimum value in this iteration, then p_p＝x_iIf f (x) is present_i)<f(p_g) Then p is_g＝x_iWherein the fitness function f is

In the formula:

is the velocity of particle i in the (k + 1) th iteration,

representing the position of particle i at the kth iteration,

is the position of the particle i in the (k + 1) th iteration, w is the weight, r₁，r₂Is a random number, p_pdIs the locally optimal solution, p, for the particle of this iteration_gdRepresenting the global optimum of the particle for this iteration, (x)_j，y_j) Two-dimensional coordinates representing a jth point of a particle;

finally, entering next iteration and repeating the formulas (1), (2) and (3), and updating the local optimal solution p_pAnd global maximum solution p_gAnd until the iteration times reach the maximum or meet the precision requirement.

Images