Improved raccoon optimization algorithm based on multiple strategies

Abstract

The invention discloses an improved raccoon optimization algorithm based on multiple strategies, and aims to solve the problems of low convergence speed, easy falling into local optimum and the like when some high-dimensional complex optimization problems are solved. Based on a traditional raccoon algorithm, the multi-strategy improved raccoon optimization algorithm provided by the invention introduces a chaotic sequence mechanism to initialize the position, introduces an adaptive nonlinear inertia weight factor in an exploration stage, introduces adaptive t distribution variation and a sparrow alerter mechanism in a development stage, and the like. And the global search capability and the convergence speed of the algorithm are effectively improved. According to the improved raccoon optimization algorithm based on multiple strategies, an effective optimization scheme is provided for the complexity problems of function optimization, engineering design optimization, machine learning parameter optimization and the like.

Description

Technical Field

[0001] The invention relates to the fields of computer science and artificial intelligence, and in particular to an improved raccoon optimization algorithm based on multiple strategies. Background Art

[0002] The Raccoon Optimization Algorithm (COA) is an intelligent optimization algorithm based on animal behavior simulation, inspired by the foraging behavior of raccoons. Raccoons show a high degree of flexibility and intelligence in the process of foraging, which makes this algorithm perform well in solving complex optimization problems. Although COA performs well in solving some complex optimization problems, it still has certain shortcomings and limitations. The convergence speed of COA is relatively slow. In the high-dimensional search space, the algorithm requires a large number of iterations to approach the global optimal solution, which increases the computational cost and time consumption; COA is prone to fall into the local optimal solution. When faced with multi-peak optimization problems, the algorithm may focus on the local area too early, resulting in difficulties in finding the global optimal solution; COA is highly sensitive to parameter settings, and different problems require different parameter adjustments, which increases the complexity of the algorithm and the difficulty of use; COA has poor adaptability in dynamic environments. Its main design is based on static environments, and it is difficult to respond quickly to environmental changes, which makes it perform poorly in real-time optimization problems. Summary of the invention

[0003] In view of the defects existing in the prior art, an intelligent optimization algorithm based on animal behavior simulation is provided. In order to achieve the above-mentioned purpose, the present invention provides the following technical solutions:

[0004] An improved raccoon optimization algorithm based on multiple strategies is characterized by comprising the following steps:

[0005] (1) Initialize the population position using chaotic mapping;

[0006] (2) Calculate the fitness value of individuals in the population;

[0007] (3) Update the optimal individual position of the population;

[0008] (4) In the exploration phase, the first half of the population introduces a nonlinear inertia weight factor to calculate and update the individual positions and fitness values ​​of raccoons; the second half of the population calculates and updates the positions of iguanas and raccoons;

[0009] (5) In the development phase, the first half of the population introduces an adaptive t-distribution mutation strategy to calculate and update the individual positions and fitness values ​​of raccoons, and the second half of the population introduces a sparrow sentinel mechanism to update the positions and fitness values ​​of raccoons;

[0010] (6) Determine whether the number of iterations has reached the maximum. If so, output the global optimal position and optimal fitness value; if not, return to the above steps to continue searching.

[0011] Preferably, in step (1), the chaotic map is a Tent map, and the formula is:

[0012] x i = l b +(u b -l b ) i

[0013]

[0014] where x i is the position of the i-th raccoon, l b is the lower limit of the variable boundary, u b is the upper limit of the variable boundary, z i is a chaotic sequence, α is a constant, which is 0.5 in the present invention;

[0015] Preferably, in step (4), the adaptive nonlinear inertia weight factor formula is:

[0016]

[0017] Among them, t is the current iteration number and T is the total iteration number.

[0018] Preferably, in step (5), the degrees of freedom in the adaptive t distribution mutation strategy are:

[0019]

[0020] Preferably, in step (5), a sparrow alert mechanism is introduced, and its formula is:

[0021]

[0022] Among them, is the j-dimensional position of the ith individual in the current iteration, x best is the current global optimal position. β is the step size control parameter and is a random number that follows a normal distribution with a mean of 0 and a variance of 1. k∈[-1, 1] is a random number, f i is the fitness value of the current raccoon individual. g and f w are the current global best and worst fitness values ​​respectively. ε is a constant, which is 10 in this invention. -50 , to avoid zero in the denominator.

[0023] In the above technical solution, the improved raccoon optimization algorithm based on multiple strategies provided by the present invention has the following beneficial effects:

[0024] 1. This invention aims to solve the problem of poor and uneven quality of the initial solution of the basic raccoon optimization algorithm by introducing a chaotic sequence mechanism to initialize the position, making the initial solution position distribution more uniform, generating high-quality initial solutions, and increasing population richness.

[0025] 2. This invention introduces a nonlinear inertia weight factor in the exploration phase to coordinate the local optimization capability and global search capability of the raccoon algorithm.

[0026] 3. This invention,introduces adaptive t-distribution variation in the development phase,to increase the diversity of the population of individuals with low fitness values,and improve the ability of the algorithm to jump out of the local optimum.

[0027] 4. This invention introduces a raccoon sentinel update mechanism during the development phase to enhance the alert capability of the raccoon algorithm, enabling it to search within an optional range. When raccoons become aware of danger, raccoons at the edge of the group will quickly move to a safe area to obtain a better position, while raccoons in the middle of the population will move randomly to get closer to other raccoons. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0029] Figure 1 A schematic diagram of a process flow of an improved raccoon optimization algorithm based on multiple strategies provided by an embodiment of the present invention;

[0030] Figure 2 The CEC2017 test set provided for the embodiment of the present invention is in the F5, F7, F9, and F10 test diagrams;

[0031] Figure 3 The CEC2021 test set provided for the embodiment of the present invention is in the F5, F7, F9, and F10 test diagrams;

[0032] Figure 4 A schematic diagram of ICEEMDA N parameter optimization for anchor cable nondestructive testing reflection echo signals provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0033] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0034] An improved raccoon optimization algorithm based on multiple strategies is characterized by comprising the following steps:

[0035] (1) Initialize the population position using chaotic mapping;

[0036] (2) Calculate the fitness value of individuals in the population;

[0037] (3) Update the optimal individual position of the population;

[0038] (4) In the exploration phase, the first half of the population introduces a nonlinear inertia weight factor to calculate and update the individual positions and fitness values ​​of raccoons; the second half of the population calculates and updates the positions of iguanas and raccoons;

[0039] (5) In the development phase, the first half of the population introduces an adaptive t-distribution mutation strategy to calculate and update the individual positions and fitness values ​​of raccoons, and the second half of the population introduces a sparrow sentinel mechanism to update the positions and fitness values ​​of raccoons;

[0040] (6) Determine whether the number of iterations has reached the maximum. If so, output the global optimal position and optimal fitness value; if not, return to the above steps to continue searching.

[0041] In the present invention, in the specific operation process, the chaotic mapping in step (1) is a Tent mapping. The Tent mapping is introduced in the initialization process of the improved raccoon algorithm to increase the ergodicity of the initial population. The formula is:

[0042] x i = l b +(u b -l b ) i

[0043]

[0044] where x i is the position of the i-th raccoon, l b is the lower limit of the variable boundary, u b is the upper limit of the variable boundary, z i is a chaotic sequence, α is a constant, which is 0.5 in the present invention;

[0045] In the present invention, in step (4), since the update of the individual raccoon position is closely related to the current position of the raccoon, an adaptive nonlinear inertia weight factor is used to adjust the correlation between the raccoon position update and the current raccoon position information. The nonlinear inertia weight factor is calculated as follows:

[0046]

[0047] Where t is the current iteration number and T is the total iteration number. The raccoon position update formula is:

[0048]

[0049] The iguana's position is updated to:

[0050]

[0051] Among them, is the j-dimensional position of the ith individual in the current iteration, x best is the current global optimal position, is the position of the i-th individual in the current iteration of the iguana, r, I and, are random numbers, f i is the fitness value of the current raccoon individual, f l is the current fitness value of the iguana, is the position after falling to the ground. In the early stage of algorithm iteration, the value of ω is small, and the update of the optimization individual position is less affected by the current raccoon position, which is conducive to the algorithm to search in a larger range and improve the global development ability of the algorithm; as the optimization process progresses, the value of ω gradually increases, and the update of the optimization individual position is more affected by the current raccoon position, which narrows the optimization range of the algorithm, helps the algorithm to search for the optimal solution, not only improves the local exploration ability of the algorithm, but also improves the convergence speed of the algorithm.

[0052] In the present invention, the degrees of freedom in the adaptive t distribution variation strategy in step (5) are:

[0053]

[0054] in, is the j-dimensional position of the ith individual in the current iteration, xbest is the current global optimal position, f(t) is a t-distributed random number, l b is the lower limit of the variable boundary, u b is the upper limit of the variable boundary, f i is the fitness value of the current raccoon individual, f AVG is the average fitness value, and r is a random number. The degree of freedom of the distribution freen increases with the number of iterations, and its numerical distribution gradually changes from Cauchy distribution to Gaussian distribution. In the early stage of the algorithm, the t distribution is Cauchy distribution, and the peak at t = 0 is smaller than that of Gaussian distribution, which is more suitable for global search. In the later stage of the algorithm, as t increases, the t distribution gradually approaches the Gaussian distribution, and the area near t = 0 is larger, which is conducive to the algorithm to solve in the current field and improve the local search ability of the later stage of the algorithm.

[0055] In the present invention, in step (5), a sparrow alert mechanism is introduced, and its formula is:

[0056]

[0057] Among them, is the j-dimensional position of the ith individual in the current iteration, x best is the current global optimal position. β is the step size control parameter and is a random number that follows a normal distribution with a mean of 0 and a variance of 1. k∈[-1, 1] is a random number, f i is the fitness value of the current raccoon individual. g and f w are the current global best and worst fitness values ​​respectively. ε is a constant, which is 10 in this invention. -50 , to avoid zero in the denominator. i >f g This means that the raccoon is on the edge of the group and is vulnerable to attacks by natural enemies. i =f g This indicates that the raccoons in the middle of the population are aware of the danger and need to get closer to other raccoons to avoid being preyed upon. k represents the direction in which the raccoons move and is also the step length control parameter.

[0058] In the present invention, Figure 2 , Figure 3 and Figure 4 These are schematic diagrams of ICEEMDAN parameter optimization on the F5, F7, F9, and F10 test graphs in the Cec2017 test set, the F5, F7, F9, and F10 test graphs in the Cec2021 test set, and the reflection echo signal of anchor non-destructive testing. Compared with other algorithms, an improved raccoon optimization algorithm based on multiple strategies enhances global and local search capabilities through multi-strategy fusion and adaptive mutation, and has higher search efficiency, adaptability, and robustness.

[0059] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. An improved raccoon optimization algorithm based on multiple strategies, characterized in that The following steps are involved: (1) Initialize the population position using chaotic mapping; (2) Calculate the fitness value of individuals in the population; (3) Update the optimal individual position of the population; (4) In the exploration phase, the first half of the population introduces a nonlinear inertia weight factor to calculate and update the individual positions and fitness values ​​of raccoons; the second half of the population calculates and updates the positions of iguanas and raccoons; (5) In the development phase, the first half of the population introduces an adaptive t-distribution mutation strategy to calculate and update the individual positions and fitness values ​​of raccoons, and the second half of the population introduces a sparrow sentinel mechanism to update the positions and fitness values ​​of raccoons; (6) Determine whether the number of iterations has reached the maximum. If so, output the global optimal position and optimal fitness value; if not, return to the above steps to continue searching.

2. The improved raccoon optimization algorithm based on multiple strategies according to claim 1 is characterized in that: In step (1), the chaotic map is the Tent map, and the formula is: x i =l b +(u b -l b )z i where x i is the position of the i-th raccoon, l b is the lower limit of the variable boundary, u b is the upper limit of the variable boundary, z i is a chaotic sequence, α is a constant, and the value in the present invention is 0.

5.

3. The improved raccoon optimization algorithm based on multiple strategies according to claim 1 is characterized in that: In step (4), the formula for the adaptive nonlinear inertia weight factor is: Among them, t is the current iteration number and T is the total iteration number.

4. The improved raccoon optimization algorithm based on multiple strategies according to claim 1 is characterized in that: In step (5), the degrees of freedom in the adaptive t-distribution mutation strategy are:

5. The improved raccoon optimization algorithm based on multiple strategies according to claim 1 is characterized in that: In step (5), the sparrow sentinel mechanism is introduced, and its formula is: Among them, is the j-dimensional position of the ith individual in the current iteration, x best is the current global optimal position. β, as the step size control parameter, is a random number that follows a normal distribution with a mean of 0 and a variance of 1. k∈[-1, 1] is a random number, f i is the fitness value of the current raccoon individual. g and f w are the current global best and worst fitness values ​​respectively. ε is a constant, which is 10 in this invention. -50 , to avoid zero in the denominator.

Images