Smell source positioning research method based on spiral flight and self-adaptive fruit fly mechanism

Abstract

The invention relates to the technical field of dangerous odor source leakage positioning, and discloses an odor source positioning research method based on spiral flight and a self-adaptive fruit fly mechanism, comprising the following steps: S1, initializing various parameters and population number; s2, using tent chaotic mapping to initialize a fruit fly group position and a current fruit fly individual concentration smelli; the smell source positioning research method based on the spiral flight and the self-adaptive fruit fly mechanism is based on a fruit fly optimization algorithm, an advanced learning strategy is fused, the speed and accuracy of a mobile robot for positioning a smell source are improved, through the new strategy, the robot can quickly respond to potential gas leakage and accurately position a leakage source, and the safety of the robot is improved. Therefore, measures are taken in time, possible fire or explosion accidents are prevented, personnel safety is protected, economic losses are reduced, and the safety and reliability of industrial production are greatly improved.

Description

Technical Field

[0001] The invention relates to the technical field of dangerous odor source leakage positioning, and in particular to a research method for odor source positioning based on spiral flight and adaptive fruit fly mechanism. Background Art

[0002] In emergency situations where toxic or hazardous chemicals leak, quickly and accurately identifying the source of the leak is crucial to reducing casualties and property losses. Therefore, how to efficiently locate the source of gas leaks has become an urgent issue to ensure public safety and promote economic development.

[0003] At present, gas leak detection in chemical plants mainly relies on traditional means, such as manual inspection and wireless sensor networks. These methods have certain limitations. Although manual inspection is intuitive, it is costly and easily affected by human factors (such as fatigue and negligence), resulting in unstable detection results. Especially in hard-to-reach areas, manual inspection often cannot effectively cover them and may also pose safety risks. Although wireless sensor networks can perform distributed monitoring, they also face problems such as high cost, environmental interference and maintenance difficulties. Especially in large-scale monitoring environments, the deployment and data processing of sensors increase the system burden and cause data security issues. At present, the research on intelligent inspection robots in gas leak detection has made certain progress in technology, mainly focusing on the optimization of odor source perception and search algorithms. For example, some researchers have proposed a multimodal gas detection method based on image recognition and sensor data fusion to improve the recognition accuracy and positioning accuracy of odor sources. In addition, some Scholars have also combined reinforcement learning and path planning algorithms to enable robots to autonomously find leak sources and achieve efficient positioning in complex environments. These technologies enable robots to not only perform simple inspection tasks, but also quickly adapt to different environmental conditions and accurately respond to gas leaks. Despite this, existing technologies still face certain challenges in the accuracy, speed and stability of odor source perception and search algorithms. Most current gas detection algorithms focus on high-precision perception of local areas and lack overall planning and optimization of the global environment. In addition, how intelligent robots maintain efficient positioning and navigation capabilities under dynamically changing environmental conditions remains a research problem. Therefore, how to design a more efficient and accurate odor source positioning algorithm, especially in complex and changeable environments, is still a key issue that needs to be urgently addressed in the current field of gas leak detection. To this end, we propose a research method for odor source positioning based on spiral flight and adaptive fruit fly mechanisms. Summary of the invention

[0004] The present invention provides a research method for odor source positioning based on spiral flight and adaptive fruit fly mechanism, which has the advantages of improving the speed and accuracy of mobile robot positioning odor source and solving the limitations of traditional fruit fly optimization algorithm in odor source positioning, and solves the problems raised by the above-mentioned background technology.

[0005] The present invention provides the following technical solution: a method for studying odor source localization based on spiral flight and adaptive fruit fly mechanism, comprising the following steps:

[0006] Step S1: Initialize various parameters and population numbers;

[0007] Step S2: Use tent chaotic mapping to initialize the position of the fruit fly group and the current fruit fly individual concentration smelli;

[0008] Step S3: introducing adaptive fruit flies into the population according to the group position, and using adaptive weights for the adaptive fruit flies;

[0009] Step S4: determining the position of the adaptive fruit fly through the Levy flight strategy and the variable spiral search strategy, and updating its position;

[0010] Step S5: retain the individual position and concentration, and the group collectively moves to the optimal individual position;

[0011] Step S6: Check whether the following conditions are met: the global optimal concentration is continuously greater than the set initial concentration threshold; if so, stop the iteration process; if not, execute S2-S5 in a loop.

[0012] As a preferred technical solution of the present invention, steps S1-S3 specifically include: the method includes tent chaotic mapping and fruit fly optimization algorithm (FOA), wherein the specific formula of tent chaotic mapping is:

[0013]

[0014] The fruit fly optimization algorithm (FOA) specifically includes a fruit fly positioning algorithm, an adaptive weight calculation algorithm, and a dynamic adaptive weight adjustment algorithm in the later stage of the algorithm:

[0015]

[0016] As a preferred technical solution of the present invention, step S4 specifically includes: the algorithm is composed of an adaptive weighted strategy algorithm and a Levy flight strategy algorithm, wherein the updated algorithm formula of the Levy flight strategy combined with the adaptive weighted strategy is:

[0017]

[0018] Where xi(t) represents the position of the ith individual at the tth iteration, ⊕ is an arithmetic symbol representing point-to-point multiplication, l is a step size control parameter, and the formula for calculating the step size is as follows:

[0019]

[0020] As a preferred technical solution of the present invention, step S5 specifically includes: the method also includes a variable spiral search strategy, the formula of the strategy is as follows:

[0021]

[0022] Where z is the dynamically changing spiral search factor. The odor source localization strategy based on spiral flight and adaptive fruit fly mechanism includes three stages: smoke plume detection, smoke plume tracking, and odor source localization:

[0023] (1) Discovering the smoke plume: First, parameter initialization is required, including: the initial positions of the robot group are Xgroup and Ygroup, the population size is n, the maximum number of iterations is Gmax, and the initial concentration threshold is C0. After initializing the initial position, the positions of the individual robots are randomly assigned according to the tent chaos map:

[0024] The specific algorithm is outlined in Algorithm 1:

[0025] Algorithm 1: Finding the smoke plume Pseudocode:

[0026]

[0027] (2) Smoke plume tracking: In the iterative optimization process, the Adaptive Spiral Flying Fruit Fly Algorithm (ASFFOA) is used to track the odor. The parameter iCount is set to record the number of times Cbest is continuously greater than Cmax. The process is shown in Algorithm 2 in the following table:

[0028] Algorithm 2: Pseudocode for plume tracking:

[0029]

[0030]

[0031] (3) Odor source location: In the odor source confirmation stage, the location is determined to be the real odor source by checking whether the odor is continuous. The real odor source neighborhood will maintain a high concentration for a long time. Therefore, a condition is set to determine whether the optimal value Cbest can be continuously greater than the set maximum concentration value Cmax. The pseudo code is as follows:

[0032] Algorithm 3: Odor source localization pseudo code:

[0033]

[0034] The present invention has the following beneficial effects:

[0035] 1. This research method for odor source localization based on spiral flight and adaptive fruit fly mechanism is based on the fruit fly optimization algorithm and integrates advanced learning strategies. It aims to improve the speed and accuracy of mobile robots in locating odor sources. Through this new strategy, the robot can quickly respond to potential gas leaks and accurately locate the source of the leak, so as to take timely measures to prevent possible fire or explosion accidents, protect personnel safety and reduce economic losses. The realization of this goal will greatly improve the safety and reliability of industrial production.

[0036] 2. The research method of odor source localization based on spiral flight and adaptive fruit fly mechanism improves the performance of the robot operation algorithm using this method by adding Levy flight strategy and adaptive weight strategy, especially solving the limitations of traditional fruit fly optimization algorithm in odor source localization, such as easy to fall into saturated convergence and low optimization accuracy, enhancing the global search capability of the algorithm and avoiding premature convergence. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is the overall flow chart of the present invention;

[0038] Figure 2 This is the distribution diagram of Levi's flight trajectory in the experimental example of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] See also Figure 1-Figure 2 , a method for odor source localization based on spiral flight and adaptive fruit fly mechanism, comprising the following steps:

[0041] Step S1: Initialize various parameters and population numbers;

[0042] Step S2: Use tent chaotic mapping to initialize the position of the fruit fly group and the current fruit fly individual concentration smelli;

[0043] Step S3: introducing adaptive fruit flies into the population according to the group position, and using adaptive weights for the adaptive fruit flies;

[0044] Step S4: determining the position of the adaptive fruit fly through the Levy flight strategy and the variable spiral search strategy, and updating its position;

[0045] Step S5: retain the individual position and concentration, and the group collectively moves to the optimal individual position;

[0046] Step S6: Check whether the following conditions are met: the global optimal concentration is continuously greater than the set initial concentration threshold; if so, stop the iteration process; if not, execute S2-S5 in a loop.

[0047] In a preferred embodiment, steps S1-S3 specifically include: the method includes tent chaotic mapping and fruit fly optimization algorithm (FOA), wherein the specific formula of tent chaotic mapping is:

[0048]

[0049] In the formula, N represents the number of particles in the chaotic sequence. According to the characteristics of tent chaotic mapping, the sequence process of generating chaos in the feasible domain is as follows:

[0050] (1) Randomly generate an initial value z0 in (0,1) and set i = 1;

[0051] (2) Iterate using the formula to generate a z sequence, and i increases by 1;

[0052] (3) If the number of iterations reaches the maximum value, stop and store the generated z sequence.

[0053] The fruit fly optimization algorithm (FOA) specifically includes the fruit fly positioning algorithm, the adaptive weight calculation algorithm, and the dynamic adaptive weight adjustment algorithm in the later stage of the algorithm:

[0054]

[0055] In order to improve the performance of the Fruit Fly Optimization Algorithm (FOA), an inertia weight w that changes dynamically with the number of iterations is introduced in the plume discovery stage. This strategy plays a key role in the early stages of the algorithm. It effectively reduces the uncertainty brought by random initialization and cooperates with the Levy flight mechanism to achieve a balance. In this way, it not only enhances the algorithm's fine search ability in local areas, but also improves its exploration efficiency on a global scale. Specifically, the dynamic adjustment mechanism of the inertia weight w enables the algorithm to respond to different optimization requirements more flexibly during the search process, thereby achieving a dynamic balance between global and local searches, providing more opportunities for finding the optimal solution. The guide fruit fly guides other individuals in the population to find food, so the introduction of adaptive weights improves the quality of a single position, enables other individuals to converge to the best position faster, and accelerates the convergence speed overall. In the second formula, w is the weight, which shows a nonlinear change characteristic in the interval [0,1]. The weight value is small in the early stage of the algorithm, which speeds up the optimization speed; in the later stage of the algorithm, the weight value gradually increases, but its change rate slows down, thus achieving a balance in the convergence performance of the algorithm. By introducing adaptive weights to dynamically adjust the position change of the fruit fly, different guidance modes are provided for the guide fruit fly at different times, which makes the search process of the algorithm more flexible. As the number of iterations increases, the fruit flies gradually converge to the optimal position. Larger weights enable individuals to move faster, thereby accelerating the convergence speed of the algorithm. This dynamic adjustment mechanism not only improves the ability of the guide fruit fly to guide the population search, but also helps the entire algorithm to achieve a good balance between global exploration and local development, thereby improving the overall performance and stability of the algorithm.

[0056] In a preferred embodiment, step S4 specifically includes: the algorithm is composed of an adaptive weighted strategy algorithm and a Levy flight strategy algorithm, wherein the updated algorithm formula of the Levy flight strategy combined with the adaptive weighted strategy is:

[0057]

[0058] Where xi(t) represents the position of the ith individual at the tth iteration, ⊕ is an arithmetic symbol representing point-to-point multiplication, l is a step size control parameter, and the formula for calculating the step size is as follows:

[0059]

[0060] In a complex industrial environment, factors such as airflow turbulence, obstacle interference, and uneven gas concentration distribution have greatly increased the difficulty of locating the odor source and affected the accuracy of positioning. In this case, it is obviously not enough to rely solely on the introduction of an adaptive weighted strategy to improve the convergence effect of the algorithm. Although the adaptive weighted strategy can accelerate the convergence of the algorithm to a certain extent, when faced with complex and changeable industrial scenarios, the algorithm still has the risk of falling into a local optimal solution. In order to further improve the global search capability of the algorithm and the diversity of solutions, the present invention introduces the Levy flight strategy, which is a random walk method with a long-tail distribution. It can Long-distance jumps are made in the search space to increase the randomness of the algorithm solution. This strategy not only enriches the diversity of population positions, but also effectively helps the algorithm jump out of the local optimal solution and avoids the algorithm from getting stuck in a complex industrial environment. By combining the adaptive weighted strategy and the Levy flight strategy, the algorithm maintains rapid convergence while also having stronger global exploration capabilities and higher operating efficiency, thereby achieving more accurate and reliable odor source positioning in a complex industrial environment. The introduction of the Levy flight strategy brings significant flexibility to the fruit fly optimization algorithm, enabling it to guide other individuals to break free from the limitations of local extreme values ​​and explore better solution space positions. The clever combination of the Levy flight mechanism and the adaptive weight not only achieves a balance in the search strategy, but also improves the quality of the solution obtained, thereby greatly enhancing the algorithm's search efficiency. This fusion strategy enables the algorithm to achieve a better balance between global exploration and local development, effectively reducing the risk of falling into the local optimum, and improving the algorithm's ability to find the global optimal solution, providing a more powerful tool for solving complex optimization problems.

[0061] In a preferred embodiment, step S5 specifically includes: the method also includes a variable spiral search strategy, the formula of the strategy is as follows:

[0062]

[0063] Where z is the dynamically changing spiral search factor. The odor source localization strategy based on spiral flight and adaptive fruit fly mechanism includes three stages: smoke plume detection, smoke plume tracking, and odor source localization:

[0064] (1) Discovering the smoke plume: First, parameter initialization is required, including: the initial positions of the robot group are Xgroup and Ygroup, the population size is n, the maximum number of iterations is Gmax, and the initial concentration threshold is C0. After initializing the initial position, the positions of the individual robots are randomly assigned according to the tent chaos map:

[0065] The specific algorithm is outlined in Algorithm 1:

[0066] Algorithm 1: Finding the smoke plume Pseudocode:

[0067]

[0068]

[0069] (2) Smoke plume tracking: In the iterative optimization process, the Adaptive Spiral Flying Fruit Fly Algorithm (ASFFOA) is used to track the odor. The parameter iCount is set to record the number of times Cbest is continuously greater than Cmax. The process is shown in Algorithm 2 in the following table:

[0070] Algorithm 2: Pseudocode for plume tracking:

[0071]

[0072]

[0073] (3) Odor source location: In the odor source confirmation stage, the location is determined to be the real odor source by checking whether the odor is continuous. The real odor source neighborhood will maintain a high concentration for a long time. Therefore, a condition is set to determine whether the optimal value Cbest can be continuously greater than the set maximum concentration value Cmax. The pseudo code is as follows:

[0074] Algorithm 3: Odor source localization pseudo code:

[0075]

[0076] On the basis of traditional spiral search, by introducing a dynamically changing spiral search factor, the search efficiency of the algorithm can be significantly improved. Specifically, in the follower position update stage, a variable spiral position update strategy is adopted to make the follower position update more flexible and changeable, thereby achieving a better balance between global search and local search. In the above formula, z is a dynamically changing spiral search factor, and its value is dynamically adjusted according to the number of iterations and specific functions to achieve flexibility and diversity in the search process. This dynamic adjustment mechanism enables the algorithm to explore the solution space more effectively during the search process and avoid falling into local optimality. At the same time, fine-tuning is used to improve the accuracy of the solution in the later stage of iteration, thereby achieving a better balance between global search and local search, significantly improving the overall performance and search efficiency of the algorithm.

[0077] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0078] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for odor source localization based on spiral flight and adaptive fruit fly mechanism, characterized in that: The following steps are involved: Step S1: Initialize various parameters and population numbers; Step S2: Use tent chaotic mapping to initialize the position of the fruit fly group and the current fruit fly individual concentration smelli; Step S3: introducing adaptive fruit flies into the population according to the group position, and using adaptive weights for the adaptive fruit flies; Step S4: determining the position of the adaptive fruit fly through the Levy flight strategy and the variable spiral search strategy, and updating its position; Step S5: retain the individual position and concentration, and the group collectively moves to the optimal individual position; Step S6: Check whether the following conditions are met: the global optimal concentration is continuously greater than the set initial concentration threshold; if so, stop the iteration process; If not satisfied, execute S2-S5 in a loop.

2. The method for odor source localization based on spiral flight and adaptive fruit fly mechanism according to claim 1, characterized in that: Steps S1-S3 specifically include: the method includes tent chaotic mapping and fruit fly optimization algorithm (FOA), wherein the specific formula of tent chaotic mapping is: The fruit fly optimization algorithm (FOA) specifically includes a fruit fly positioning algorithm, an adaptive weight calculation algorithm, and a dynamic adaptive weight adjustment algorithm in the later stage of the algorithm:

3. The method for odor source localization based on spiral flight and adaptive fruit fly mechanism according to claim 1, characterized in that: Step S4 specifically includes: the algorithm is composed of an adaptive weighted strategy algorithm and a Levy flight strategy algorithm, wherein the updated algorithm formula of the Levy flight strategy combined with the adaptive weighted strategy is: Where xi(t) represents the position of the ith individual at the tth iteration, ⊕ is an arithmetic symbol representing point-to-point multiplication, l is a step size control parameter, and the formula for calculating the step size is as follows:

4. The method for locating odor sources based on spiral flight and adaptive fruit fly mechanism according to claim 1, characterized in that: Step S5 specifically includes: the method also includes a variable spiral search strategy, the formula of the strategy is as follows: Where z is the dynamically changing spiral search factor. The odor source localization strategy based on spiral flight and adaptive fruit fly mechanism includes three stages: smoke plume detection, smoke plume tracking, and odor source localization: (1) Discovering the smoke plume: First, parameter initialization is required, including: the initial positions of the robot group are Xgroup and Ygroup, the population size n, the maximum number of iterations Gmax, and the initial concentration threshold C0. After initializing the initial positions, the positions of the individual robots are randomly assigned according to the tent chaotic map: The specific algorithm is outlined in Algorithm 1: Algorithm 1: Finding the smoke plume Pseudocode: Initialization parameters Randomly assign the positions of individual robots according to the tent chaos map do{ for(i=1;i <n;i++){ Keep recording Ci and Index / / Ci is the concentration of the i-th robot individual} for(i=1;i <n;i++){Cbest=max(Ci)} / / Find the maximum population concentration value Cbest if Cbest>C0 then goto{smoke plume tracking phase}} until G>Gmax / / Find the maximum population concentration value Cbest / / G represents the current number of iterations (2) Smoke plume tracking: In the iterative optimization process, the Adaptive Spiral Flying Fruit Fly Algorithm (ASFFOA) is used to track the odor. The parameter iCount is set to record the number of times Cbest is continuously greater than Cmax. The process is shown in Algorithm 2 in the following table: Algorithm 2: Pseudocode for plume tracking: (3) Odor source location: In the odor source confirmation stage, the location is determined to be the real odor source by checking whether the odor is continuous. The real odor source neighborhood will maintain a high concentration for a long time. Therefore, a condition is set to determine whether the optimal value Cbest can be continuously greater than the set maximum concentration value Cmax. The pseudo code is as follows: Algorithm 3: Odor source localization pseudo code: if(G≤Gmax and iCount=d) / / G is the current number of iterations then{declare that the task of locating the source of the odor is successful} else if(G=Gmax and iCount <d) then{claim that the task of locating the source of the odor has failed} else if(G <Gmax and iCount<d) then goto{smoke plume tracking phase}end.

Images