Indoor toxic and harmful gas multi-potential source positioning method and system

Abstract

The invention discloses an indoor toxic and harmful gas multi-potential source positioning method and system, and the method comprises the steps: carrying out the traceability updating of a measurement point location in a modeling space, introducing a dynamic class-dividing strategy to carry out the individual class-dividing of a population in the space based on a teaching and learning optimization algorithm, carrying out the teaching and learning stage of the individual of each class to update the position of the individual, and carrying out the positioning of the measurement point location. Teaching factors are dynamically adjusted to change teaching paths, individual values are designed, learning paths are changed, elite preservation screening is carried out on individuals, a contact person is introduced to serve as a global planning role for individual injection, the worst individual in a population is replaced, and the updating position of each measurement point position is determined; performing spatial modeling on the indoor environment, receiving an updated position and guiding the measurement point to move to the updated position; the concentration of indoor poisonous and harmful gas detected at the measuring point is fed back to the point tracing unit for subsequent position updating, so that the success rate and the positioning precision of multi-gas source tracing are improved, and the defects in the field of multi-source positioning in the current indoor environment are overcome.

Description

Technical Field

[0001] The present invention belongs to the technical field of pollution source positioning, and in particular relates to a method and system for positioning multiple potential sources of indoor toxic and harmful gases. Background Art

[0002] Toxic and hazardous gases pose a serious threat to human health, being highly irritating and potentially carcinogenic. The severity of these hazards increases with prolonged exposure. As more people spend their time indoors, indoor toxic and hazardous gas pollution has become a global public health concern. The potential sources of indoor pollution are complex, including formaldehyde, benzene, and other indoor decorative pollutants. Therefore, rapid and accurate identification of multiple potential sources is essential for effective source control and refined governance.

[0003] Current methods for tracing the source of toxic and hazardous gases indoors include fixed sensor models that use multiple sensor arrays combined with models to locate pollution sources. However, this method requires pre-modeling of the environment, lacking effectiveness, and is highly dependent on installed sensor arrays, making it inconvenient. There are also active tracing methods based on mobile platforms, but these algorithms, such as PSO, IAFSA, and RRT, suffer from premature convergence, high parameter complexity, or computational complexity, and lack the ability to locate multiple potential sources. Furthermore, the instruments used in current tracing methods are predominantly electrochemical sensors, which suffer from poor long-term stability, drift, and poor interference resistance, such as temperature drift. They are also susceptible to airflow, resulting in falsely high readings from air conditioning and exhaled breath. The presence of other volatile organic compounds can lead to false positives of significantly exceeded standards. Complex indoor environments often contain multiple potential volatile sources, with complex compositions and a wide range of sources, posing significant challenges to the development of rapid and accurate tracing methods.

[0004] The teaching-learning optimization algorithm has the advantages of low algorithm parameter dependence, fast convergence, and easy implementation. It has great potential application scenarios in the tracing of indoor toxic and harmful gases. Based on the principles of the teaching-learning optimization algorithm, by improving teaching factors, balancing exploration and development, dynamic class division and other strategy optimizations, it is possible to break through the limitations of indoor multi-source positioning; multi-pass reflectance in-situ detection instruments based on the principle of spectroscopy detection use the characteristic absorption of different substances to simultaneously detect and distinguish multiple substances. It can break through the limitations of poor anti-interference and easy drift of readings of detection instruments such as sensors, and achieve accurate detection of toxic and harmful gases in complex indoor environments. Based on this, a multi-gas source positioning method is developed and used as an optimization strategy to guide high-sensitivity in-situ detection instruments for precise detection. It can achieve rapid and accurate identification of multiple potential sources of indoor toxic and harmful gases, which is of great value in improving the deficiencies of indoor gas multi-source positioning and refined pollution control. Summary of the Invention

[0005] In view of the above, the purpose of the present invention is to provide a method and system for locating multiple potential sources of indoor toxic and harmful gases, so as to realize the accurate identification of multiple indoor pollution sources, improve the success rate and positioning accuracy of multi-gas source tracing, and improve the shortcomings in the field of multi-source positioning in the current indoor environment.

[0006] To achieve the above-mentioned purpose of the invention, an embodiment provides a system for locating multiple potential sources of indoor toxic and harmful gases, comprising: The point traceability unit is used to trace and update the measurement points in the modeling space. Specifically, based on the teaching and learning optimization algorithm, multiple measurement points are used as individuals to initialize a population composed of individuals. A dynamic class division strategy is introduced to divide the population in space into regions to realize the class division of individuals within the population. The teaching and learning stages are carried out for individuals in each class to update the individual positions. In the teaching stage, the teaching factors are dynamically adjusted to change the teaching path. In the learning stage, the individual values ​​are designed and the learning path is changed based on the individual values. At the same time, teaching optimization is also carried out to realize the elite retention and screening of individuals to determine the updated position of each measurement point. The spatial positioning unit is used to perform spatial modeling for the indoor environment, receive the updated position sent by the point traceability unit, and guide the measurement point to move to the updated position; The detection unit is used to perform high-precision in-situ gas detection at the measurement point and feed back the detected indoor toxic and harmful gas concentration to the point traceability unit for subsequent position update.

[0007] Preferably, a dynamic class division strategy is introduced to divide the population in space into regions to realize class division of individuals within the population, including: The dynamic class division strategy is triggered according to the set class division interval, where the class division interval is set according to the number of iterations. The class division process according to the dynamic class division strategy is expressed as: in, and For two different individuals, and is the individual index, Represents the class index, Represents an individual and The distance between Indicates the t Dynamic distance threshold at iterations, represents the basic threshold, T represents the maximum number of iterations, Indicates the exponential factor, ranging from 0.01 to 0.99. Represents an individual The class you are in, Indicates that no class division will be carried out.

[0008] Preferably, dynamically adjusting the teaching factors to change the teaching path during the teaching phase includes: Dynamically adjust the teaching factor according to the number of iterations : in, is the current iteration number, T Indicates the maximum number of iterations; Based on teaching factors Change the teaching path to update the individual position as follows: in, and Respectively represent individual students in the teaching stage i From the teacher's position before and after learning, represents the position of the individual teacher, Represents a random number between 0 and 1. Represents the average position of all individual students in each class.

[0009] Preferably, individual values ​​are designed during the learning stage and the learning path is changed based on individual values, including: The design individual value is: in, Indicates the class number. Represents the individual index of students in the class, Represents individual students in class k The value of and are the minimum fitness and maximum fitness of the current class respectively, Is the current individual The fitness of is the global maximum fitness of all classes, and As the weight distribution coefficient of class teaching and individual value respectively, satisfying the constraints ; Change learning paths based on individual values, including: in, and Respectively represent individual students in the academic stage i Learning the position before and after, Represents individual students j Position before learning, and Respectively and The corresponding individual fitness, where the fitness is the indoor toxic and harmful gas concentration detected by the detection unit.

[0010] Preferably, teaching optimization is also performed to achieve elite retention and screening of individuals, including: The screening strategy based on the following formula is used to select elite individuals in the teaching optimization process: in, and Represent the new position and old position of the individual respectively. If the fitness of the new position Greater than or equal to the fitness of the old position , then the new position of the individual is retained, otherwise the old position of the individual is retained.

[0011] Preferably, a liaison is introduced as a global planning role in the teaching and learning optimization algorithm to explore uncovered areas and add diversity to the entire optimization process. The liaison's trajectory is controlled by a dynamic spiral equation and its position is updated. Based on the fitness of the new position and the individual fitness of the population, the liaison is injected into the population to replace the worst individuals in the population.

[0012] Preferably, the movement trajectory of the liaison is controlled by a dynamic spiral equation and the position is updated, including: in, The initial location for the contact search, is the maximum search radius associated with the modeling space, are the radius expansion coefficient, angle attenuation coefficient and basic rotation angle respectively, and express Second and The rotation angle at the iteration, express The radius at the iteration, represents the maximum number of iterations, ( , )express The updated position at iteration .

[0013] Preferably, the individual injection of the liaison into the population is achieved based on the fitness of the new position and the individual fitness of the population to replace the worst individual in the population, including: Based on the fitness of the new position and the individual fitness of the population, the following injection mechanism is introduced to implement the injection of individuals from the liaison to the population to replace the worst individuals in the population: in, represents the new fitness of the contact, represents the individual fitness of the population, when hour = , which means that the position of the worst individual in the population is replaced by the new position of the liaison, while the liaison also maintains its new position for the next round of position update.

[0014] Preferably, the spatial positioning unit includes a ranging module and a modeling module, wherein the ranging module is used to obtain the spatial dimensions of the indoor environment and obtain the coordinates of its own position during the traceability detection process; the modeling module is used to construct the environmental space coordinates based on the distance information of the ranging unit, determine the current position of the detection instrument, and take a real-life picture and upload it as a model base map to facilitate the combination of the spatial modeling model and the coordinates of the actual scene.

[0015] To achieve the above-mentioned purpose of the invention, an embodiment of the present invention further provides a method for locating multiple potential sources of indoor toxic and harmful gases. The method adopts the above-mentioned system for locating multiple potential sources of indoor toxic and harmful gases, and includes the following steps: Use spatial positioning units to perform spatial modeling for indoor environments; Use the point traceability unit to trace and update the measurement points in the modeling space; Use the spatial positioning unit to receive the updated position sent by the point traceability unit and guide the measurement point to move to the updated position; The detection unit is used to perform high-precision in-situ gas detection at the measurement point, and the detected indoor toxic and harmful gas concentration is fed back to the point traceability unit for subsequent location updates.

[0016] Compared with the prior art, the present invention has the following beneficial effects: (1) The multi-potential source positioning method provided by the present invention breaks through the limitations of traditional algorithms with strong parameter dependence and poor multi-source positioning effect. Through optimization such as random initialization, dynamic class division and multi-teacher setting, liaison officer global planning and teaching factor adjustment, it improves the success rate of tracing and positioning accuracy, and expands the ability to accurately locate multiple potential sources.

[0017] (2) To overcome the drawbacks of non-strategic planning and blindness of traditional indoor observation methods, a multi-potential source positioning system for indoor toxic and harmful gases based on the traceability algorithm is developed to achieve rapid and accurate identification and detection of multiple potential sources; in terms of detection instruments, highly sensitive in-situ detection instruments based on the principle of spectroscopy measurement are selected instead of sensors to improve anti-interference performance, which is suitable for accurate gas detection in complex multi-source environments indoors. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 1 is a schematic structural diagram of a multi-potential source positioning system for indoor toxic and harmful gases provided by an embodiment; Figure 2 This is a flow chart of a point tracing algorithm based on a teaching and learning optimization algorithm provided in an embodiment; Figure 3 This is a schematic diagram of the effect of point tracing provided by the embodiment; Figure 4 This is a performance parameter diagram of point traceability provided by the embodiment; Figure 5 This is a multi-source exploration capability diagram of point traceability provided by the embodiment; Figure 6 This is a flow chart of a method for locating multiple potential sources of indoor toxic and harmful gases combined with a positioning system provided in an embodiment; Figure 7 The embodiment provides a detailed flow chart of the method for locating multiple potential sources of indoor toxic and harmful gases provided by the embodiment; Figure 8 This is a detection roadmap provided by the embodiment. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and do not limit the scope of protection of the present invention.

[0021] The embodiment provides a method and system for locating multiple potential sources of indoor toxic and harmful gases, which are mainly aimed at complex indoor environments to achieve traceability detection of multiple potential sources. It takes the strategic optimization of "multi-source positioning" combined with "high-precision" detection instruments as the core, uses high-precision detection instruments to obtain in-situ measurement information and feeds it back to the point tracing algorithm for subsequent detection point updates, until it approaches the gas source position, completing the application of the multi-source positioning algorithm in the field of indoor toxic and harmful gas traceability. The advantages of the system are: 1. Applying the multi-source positioning algorithm based on teaching optimization to the field of indoor toxic and harmful gas detection, realizing the implementation of the algorithm, and making the detection of different indoor spaces more universal; 2. Breaking through the shortcomings of non-strategic planning and blindness of traditional indoor observation methods, achieving rapid and accurate identification of sources; 3. Using a multi-trans detection instrument based on the principle of spectroscopy detection on the detection instrument, the accuracy is higher than that of the sensor and the anti-interference ability is stronger.

[0022] In the embodiment, the multi-potential point tracing algorithm is converted into a version suitable for indoor gas source positioning, and then combined with a high-precision in-situ inspection instrument based on the spectroscopy detection principle to collect real and reliable concentration data, and the detection position is identified through spatial positioning, forming a system as follows: Figure 1 The system, shown in Figure 1, consists of a point-based tracing unit, a spatial positioning unit, and a detection unit. The point-based tracing algorithm provides initial detection point information and optimizes data from the detection unit to provide new detection point information. This enables strategic detection, guides the detection instrument toward the gas source, and ultimately completes the tracing task. Its multi-source detection accuracy exceeds 90%, with positioning accuracy better than 0.2m, and can be expanded to simultaneously detect at least five sources.

[0023] In this embodiment, the point traceability unit is used to update the traceability of measurement points in the modeling space using a point traceability algorithm. Individuals in the population correspond to measurement points, fitness corresponds to the detected concentration of toxic and hazardous gases, and the number of iterations corresponds to the number of optimization attempts. The point traceability algorithm initializes the detection points and optimizes them based on the data provided by the detection unit, providing new detection point information and guiding the detection instrument toward the gas source.

[0024] Research has found that the teaching and learning optimization algorithm has low parameter dependence, fast convergence, and is easy to implement. It has great potential application scenarios in the tracing of indoor toxic and harmful gases, so this algorithm is used as the basis. Therefore, the point tracing algorithm proposed in this invention is based on the teaching and learning optimization algorithm. Taking into account the objective fact that there are often multiple pollution sources in complex indoor environments, the optimization direction is clarified to adapt to the accurate identification of multiple sources. Figure 2As shown, the point tracing algorithm of the present invention is based on "dynamic class division, global exploration", "class teaching, local development" combined with "liaison officer global planning, elite injection" as the core, to solve the problems of easy falling into local convergence, balancing global exploration and local development, and low success rate and positioning accuracy of multi-source positioning.

[0025] like Figure 2 As shown in the figure, the point tracing algorithm of the measurement points includes four parts: population initialization, liaison officer setting, dynamic class division and multiple teachers, and teaching optimization.

[0026] The population initialization stage sets the initial detection route. Multiple measurement points are used as individuals to initialize a population of individuals. Complex indoor environments often have multiple potential sources of toxic and hazardous gas releases. Therefore, each measurement point has an equal probability of being a pollution source. Therefore, the initial detection coordinates should be randomly generated, and correlation between variables should be minimized. To avoid losing source information, a sampling point generation strategy based on the Latin hypercube sampling principle is used to obtain initial measurement point coordinate information, ensuring as even coverage as possible across the entire space. This process covers the parameter space more evenly than simple random sampling and is more suitable for initializing multi-source location optimization problems.

[0027] The liaison setting part performs global planning. To improve the efficiency and computational cost of multi-source localization, a smaller population size is often used. This results in the inability to fully cover the possible release source area during the population initialization phase, leading to the omission of potential sources. Setting the liaison as a global planning role can achieve exploration of the coverage area and add diversity to the entire optimization process. The liaison's motion trajectory is controlled by a dynamic spiral equation and its position is updated, including dual adjustment of radius and angle: in, The initial location for the contact search, is the maximum search radius associated with the modeling space, are the radius expansion coefficient, angle attenuation coefficient and basic rotation angle respectively, and express Second and The rotation angle at the iteration, express The radius at the iteration, represents the maximum number of iterations, ( , )express The updated position at iteration .

[0028] By setting up liaisons, we can explore the global situation and avoid falling into local optimality. The liaisons themselves do not participate in the subsequent teaching optimization and position update of the population, but instead travel along a specific path to obtain global information. However, their fitness can be injected into the population as a high-quality solution for interaction. That is, based on the fitness of the new position and the individual fitness of the population, the liaisons are injected into the population to replace the worst individuals in the population. The injection mechanism is: in, represents the new fitness of the contact, represents the individual fitness of the population, when hour = , which means that the liaison's new position replaces the worst individual in the population, achieving elite injection, replacing inferior solutions for optimization, accelerating convergence, and maintaining high-quality individuals. At the same time, the liaison also maintains its new position for the next round of position updates.

[0029] In an embodiment, when the population diversity is low, the introduced liaison can provide better solutions and even act as a teacher for teaching optimization, thereby improving population diversity and avoiding falling into local optimality; in other cases, the inferior solutions of the population can also be replaced to avoid their interference with teaching optimization and accelerate convergence.

[0030] Dynamic class division and multi-teacher part deal with the problem of multiple potential sources through class division teaching. Class division is carried out according to the spatial position density of individuals in the population, realizing regional multi-class independent teaching and maintaining the ability to explore multiple regions. After class division, each class is assigned a teacher to teach independently, and teaching optimization is carried out based on the fitness information of individuals in the population to achieve local development. Since the individual position changes after optimization, in order to avoid ineffective exploration in the passive area and reduce the algorithm cost, the dynamic class division strategy is triggered according to the set class division interval. Among them, the class division interval is set according to the number of iterations. The process of class division according to the dynamic class division strategy is expressed as: in, and For two different individuals, and is the individual index, Represents the class index, Represents an individual and The distance between Indicates the t Dynamic distance threshold at iterations, represents the basic threshold, T represents the maximum number of iterations, It represents the exponential factor, ranging from 0.01 to 0.99, with a preferred value of 0.5. Represents an individual The class you are in, Indicates that no class division will be carried out.

[0031] At this point, individuals with high population fitness tend to cluster near the gas source, making it logical to divide classes based on spatial location and density. Through this dynamic class division and multiple teacher setup, we maintain multi-regional exploration capabilities while developing local sources, maintaining diversity to adapt to the multi-source problem.

[0032] Teaching optimization simulates the knowledge transfer between teachers and students in the teaching process, combines the population fitness information, and ensures that individuals converge to the source position. It is mainly divided into the teaching stage and the learning stage. In the teaching stage, individuals with higher fitness are set as teachers to teach students. The traditional teaching optimization method is a single-teacher mode. There are multiple teachers set for optimization, but the teacher is selected as the centroid of multiple teachers. However, it is essentially still optimized for a single source. The present invention handles the multi-source problem by setting multiple teachers in different classes. At the same time, a dynamic adjustment teaching factor TF is set to change the teaching path, and TF is dynamically adjusted according to the number of iterations: in, is the current iteration number, T Indicates the maximum number of iterations; Based on teaching factors Change the teaching path to update the individual position as follows: in, and Respectively represent individual students in the teaching stage i From the teacher's position before and after learning, represents the position of the individual teacher, Represents a random number between 0 and 1. Represents the average position of all individual students in each class.

[0033] By setting the teaching factor TF, a large value of the teaching factor in the early stage of exploration promotes global exploration, while a low value in the later stage turns to local development, thereby achieving a balance between exploration and development to face multi-source problems.

[0034] During the learning phase, in order to improve the algorithm's adaptability and avoid homogenized learning for different individuals, we consider that different individuals have different values. Taking into account the impact of class division, we design optimized individual values ​​and change the learning path based on these values. The individual values ​​are: in, Indicates the class number. Represents the individual index of students in the class, Represents individual students in class k The value of and are the minimum fitness and maximum fitness of the current class respectively, Is the current individual The fitness of is the global maximum fitness of all classes, and As the weight distribution coefficient of class teaching and individual value respectively, satisfying the constraints ; Change learning paths based on individual values, including: in, and Respectively represent individual students in the academic stage i Learning the position before and after, Represents individual students j Position before learning, and Respectively and The corresponding individual fitness, where the fitness is the indoor toxic and harmful gas concentration detected by the detection unit.

[0035] The improved learning stage has higher adaptability and takes individual value into account. Individuals with high fitness have larger individual value coefficients and perform source detection, while individuals with low fitness conduct global exploration. At the same time, considering the impact of class division, the current class may have low fitness values. To avoid ineffective learning in low-value groups, the weight of class teaching is added to the individual value model. This ensures that other individuals with low fitness have the possibility of learning to higher values ​​and avoid being trapped in local optimality.

[0036] The key process in the teaching and learning phases lies in the elite retention strategy. This strategy is implemented during both the teaching and learning phases to prevent population degradation. While individual positions are updated during the teaching phase, there's a potential bias toward lower fitness. To ensure the quality of the population solution while maintaining high fitness during the learning phase to improve convergence efficiency, an elite retention strategy is implemented for the new individual positions generated during the teaching optimization process. This is achieved using the following screening strategy: in, and Represent the new position and old position of the individual respectively. If the fitness of the new position Greater than or equal to the fitness of the old position , the new position of the individual is retained, otherwise the old position of the individual is retained. The elite retention strategy avoids population "degeneration" and provides convergence guarantee.

[0037] Based on the four steps described above, namely population initialization, liaison officer setup, dynamic class division with multiple teachers, and teaching optimization, the updated position of each measurement point can be determined. The success rate and positioning accuracy were verified through 100 dual-gas source simulation experiments in a 10m x 10m environment. While maintaining a certain population size, the current algorithm achieves a multi-source detection success rate exceeding 90% and a positioning accuracy better than 0.2m, significantly outperforming traditional traceability algorithms in multi-source positioning performance.

[0038] After a teaching optimization is complete, the updated point location is sent to the spatial positioning unit to ensure that the detection unit obtains gas concentration information at the correct location. If the number of explorations has not been reached, the point concentration detection guided by the source tracing algorithm is repeated, and dynamic class division is performed based on the division criteria. At this time, individuals with high population fitness tend to be concentrated near the gas source, making class division based on spatial location density relationships logical. Through this dynamic class division and multi-teacher setup, we maintain multi-region exploration capabilities while developing local sources, maintaining population diversity to adapt to the multi-source problem.

[0039] In the embodiment, a simulation experiment is also conducted in an indoor scene under the influence of multiple air sources based on the Gaussian diffusion model to test the algorithm performance. The experimental scene is a 10×10m space with two air sources. Figure 3 The diagram shows the effect of the multi-potential source positioning method. The white dotted line is the "liaison" path, and the other broken lines represent the search paths of the individuals in the population. From the simulation results, the optimized settings of the liaison and dynamic class division enable the algorithm to achieve the goal of multi-source positioning. The performance of the method is then characterized by the success rate, convergence speed and positioning accuracy. The number of experiments is set to 100. If the optimal position is within 0.5m of the gas source, it is considered successful. The success rate is the ratio of the number of successes to the total number of experiments. The convergence speed is characterized by the average number of iterations. The positioning accuracy is measured by the distance between the optimal position and the actual gas source. In the multi-source state, the maximum positioning distance is selected. Figure 4 The performance parameter diagram of the multi-potential point tracing algorithm is shown. As the number of individuals in the population increases, the success rate of multi-source positioning also increases, which can exceed 90%. The convergence speed is improved, and the positioning task can be completed after 10 iterations. The positioning accuracy is improved, which is better than 0.2m. Continue to test the multi-source exploration capability, such as Figure 5 As shown in the figure, the horizontal and vertical coordinates of the points represent the population size and the number of iterations. The results show that the current algorithm can simultaneously locate at least five sources with a success rate exceeding 60%, ensuring the feasibility of tracing multiple indoor pollution sources.

[0040] The above-mentioned point tracing algorithm provided by the present invention first initializes the individuals in the population and divides them into classes according to their spatial positions, dividing the entire space into multiple areas for global exploration; then a teacher is assigned to each class to conduct independent teaching and conduct local development within the class. At the same time, a liaison is set up as a global plan to inject high-quality solutions into the population during the exploration process. The advantages of this point tracing algorithm are: 1. Setting dynamic class division and multi-teacher mode, and class division based on spatial density can reduce the investment cost of a single source search, balance exploration and development to accommodate multiple sources; 2. Setting a liaison as a global plan and injecting high-quality solutions into the population can increase the convergence speed of source exploration and avoid falling into local optimality; 3. Compared with traditional indoor tracing methods, this method can accurately locate multiple potential sources at the same time.

[0041] In an embodiment, the spatial positioning unit is used to perform spatial modeling for the indoor environment, receive the updated position sent by the point traceability unit, and guide the measurement point to the updated position, assisting the instrument in performing concentration detection at the corresponding position. Specifically, the spatial positioning unit includes a distance measurement module and a modeling module. The distance measurement module is used to obtain the spatial dimensions of the indoor environment based on the principle of laser ranging and obtain the coordinates of its own position during the traceability detection process. The modeling module is used to construct the environmental spatial coordinates based on the distance information of the distance measurement unit, determine the current position of the detection instrument, and take a real-life picture and upload it as a model base map to facilitate the combination of the spatial modeling model and the coordinates of the actual scene.

[0042] The detection unit performs high-precision in-situ gas detection at the measurement point and feeds the detected, accurate, high-spatial-resolution indoor toxic and hazardous gas concentrations back to the point traceability unit for subsequent location updates. Based on the principle of spectroscopy detection, it utilizes an optical cavity composed of two highly reflective mirrors to achieve long-path absorption of gases. This system offers advantages such as high detection accuracy, a wide range of detectable species, strong anti-interference capabilities, and high spatial resolution, making it suitable for indoor toxic and hazardous gas detection.

[0043] Specifically, in the detection unit, in order to achieve simultaneous detection of multiple gases, LEDs are used as broadband light sources to achieve high optical power and wide coverage fluctuation of optical signals. Multiple reflections are achieved through a light intensity system composed of high-reflectivity mirrors with a reflectivity of more than 99.9% on both sides, achieving a kilometer optical path and reducing the instrument's detection limit. Through the optimization of the iterative DOAS algorithm, the limitations of traditional technologies affected by absolute light intensity are overcome, stability is improved, and a margin is provided for the lightweight design of the instrument. Because its gas path pumps the gas at the measurement point into the cavity for detection, high spatial resolution in-situ detection can be achieved, which is suitable for complex indoor environments. The detection unit performs high-precision in-situ detection of gases at the corresponding points, obtains accurate and high spatial resolution concentration information, and feeds it back to the traceability algorithm unit for subsequent coordinate updates.

[0044] Based on the above-mentioned indoor toxic and harmful gas multi-potential source positioning system, the embodiment also provides an indoor toxic and harmful gas multi-potential source positioning method, such as Figure 6 and Figure 7 As shown, the following steps are included: S1. System initialization, including constructing spatial coordinates, initializing detection points, and initializing detection instruments.

[0045] Specifically, the distance to the indoor space to be measured is detected, a spatial coordinate model is constructed, and the current position is clarified; the traceability algorithm is executed to obtain the initial detection point information; the instrument is initialized and reference spectrum is collected to prepare for subsequent point concentration detection.

[0046] S2. Use the detection instrument to obtain the initial point concentration information and feed it back to the point traceability algorithm for teaching optimization, update the subsequent measurement point information, and guide the measurement point to move to the corresponding position through the spatial positioning unit.

[0047] Specifically, the point traceability algorithm sends the generated initial point information to the spatial positioning unit, which uses the unit to assist the instrument in performing detection at the corresponding points. Each point is detected for 10 seconds, and the coordinates and concentration information of multiple points are fed back to the point traceability algorithm. The point traceability algorithm performs dynamic class division and multi-teacher teaching optimization, and dynamically adjusts the teaching factors and individual value model based on the actual indoor conditions. It uses an elite retention strategy to maintain the high-quality solution of the population. It also introduces a liaison as a global planning role, enabling the injection of individuals into the population by the liaison to replace the worst individuals in the population. The updated coordinates are then sent to the system to continue the detection task.

[0048] S3. Repeat step S2 to obtain updated point and concentration information, and continuously approach the gas source location, thereby realizing multi-potential source positioning detection in an indoor environment.

[0049] Specifically, step S2 is repeated to record the detected point information, while continuing the tasks of concentration detection and coordinate optimization. Figure 8 The image shows the detection route map of the system during the task of locating multiple potential sources of toxic and harmful gases indoors. The clumps in the image represent potential sources. There are currently four volatile sources in the room. Each point in the image represents the detection point information under the guidance of the traceability algorithm. Due to the establishment of a liaison officer and the combination of dynamic class division and multi-teacher mode, multi-region exploration is achieved globally. It can be clearly seen from the image that the exploration routes of at least four classes are constantly approaching the gas source locations determined by each of them. Two of the sources are relatively close, and the dynamic class division strategy can also switch between different source explorations to prevent misjudgment of one source. The entire exploration process used 69 detection points, plus the experimental preparation time such as movement, a total of 20 minutes, achieving accurate identification of the four potential sources in the indoor environment.

[0050] The specific implementation methods described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above is only the most preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A multi-potential source positioning system for indoor toxic and harmful gases, characterized in that: include: The point traceability unit is used to trace and update the measurement points in the modeling space. Specifically, based on the teaching and learning optimization algorithm, multiple measurement points are used as individuals to initialize a population composed of individuals. A dynamic class division strategy is introduced to divide the population in space into regions to realize the class division of individuals within the population. The teaching and learning stages are carried out for individuals in each class to update the individual positions. In the teaching stage, the teaching factors are dynamically adjusted to change the teaching path. In the learning stage, the individual values ​​are designed and the learning path is changed based on the individual values. At the same time, teaching optimization is also carried out to realize the elite retention and screening of individuals to determine the updated position of each measurement point. The spatial positioning unit is used to perform spatial modeling for the indoor environment, receive the updated position sent by the point traceability unit, and guide the measurement point to move to the updated position; The detection unit is used to perform high-precision in-situ gas detection at the measurement point and feed back the detected indoor toxic and harmful gas concentration to the point traceability unit for subsequent position update.

2. The indoor toxic and harmful gas multi-potential source positioning system according to claim 1 is characterized in that: A dynamic class division strategy is introduced to divide the population into regions within the space and realize the individual class division within the population, including: The dynamic class division strategy is triggered according to the set class division interval, where the class division interval is set according to the number of iterations. The class division process according to the dynamic class division strategy is expressed as: in, and For two different individuals, and is the individual index, Represents the class index, Represents an individual and The distance between Indicates the t Dynamic distance threshold at iterations, represents the basic threshold, T represents the maximum number of iterations, Indicates the exponential factor, ranging from 0.01 to 0.

99. Represents an individual The class you are in, Indicates that no class division will be carried out.

3. The indoor toxic and harmful gas multi-potential source positioning system according to claim 1, characterized in that: Dynamically adjust teaching factors and change teaching paths during the teaching phase, including: Dynamically adjust the teaching factor according to the number of iterations : in, is the current iteration number, T Indicates the maximum number of iterations; Based on teaching factors Change the teaching path to update the individual position as follows: in, and Respectively represent individual students in the teaching stage i From the teacher's position before and after learning, represents the position of the individual teacher, Represents a random number between 0 and 1. Represents the average position of all individual students in each class.

4. The indoor toxic and harmful gas multi-potential source positioning system according to claim 1, characterized in that: Design individual values ​​during the learning stage and change the learning path based on individual values, including: The design individual value is: in, Indicates the class number. Represents the individual index of students in the class, Represents individual students in class k The value of and are the minimum fitness and maximum fitness of the current class respectively, Is the current individual The fitness of is the global maximum fitness of all classes, and As the weight distribution coefficient of class teaching and individual value respectively, satisfying the constraints ; Change learning paths based on individual values, including: in, and Respectively represent individual students in the academic stage i Learning the position before and after, Represents individual students j Position before learning, and Respectively and The corresponding individual fitness, where the fitness is the indoor toxic and harmful gas concentration detected by the detection unit.

5. The indoor toxic and harmful gas multi-potential source positioning system according to claim 1, characterized in that: At the same time, we also carry out teaching optimization to achieve individual elite retention screening, including: The screening strategy based on the following formula is used to select elite individuals in the teaching optimization process: in, and Represent the new position and old position of the individual respectively. If the fitness of the new position Greater than or equal to the fitness of the old position , then the new position of the individual is retained, otherwise the old position of the individual is retained.

6. The indoor toxic and harmful gas multi-potential source positioning system according to claim 1, characterized in that: In the teaching and learning optimization algorithm, a liaison is also introduced as a global planning role. The liaison's movement trajectory is controlled by the dynamic spiral equation and its position is updated. Based on the fitness of the new position and the individual fitness of the population, the liaison is injected into the population to replace the worst individual in the population.

7. The indoor toxic and harmful gas multi-potential source positioning system according to claim 6, characterized in that: The movement trajectory of the liaison is controlled by the dynamic spiral equation and the position is updated, including: in, The initial location for the contact search, is the maximum search radius associated with the modeling space, are the radius expansion coefficient, angle attenuation coefficient and basic rotation angle respectively, and express Second and The rotation angle at the iteration, express The radius at the iteration, represents the maximum number of iterations, ( , )express The updated position at the iteration.

8. The indoor toxic and harmful gas multi-potential source positioning system according to claim 6, characterized in that: Based on the fitness of the new position and the individual fitness of the population, the liaison is injected into the population to replace the worst individual in the population, including: Based on the fitness of the new position and the individual fitness of the population, the following injection mechanism is introduced to implement the injection of individuals from the liaison to the population to replace the worst individuals in the population: in, represents the new fitness of the contact, represents the individual fitness of the population, when hour = , which means that the position of the worst individual in the population is replaced by the new position of the liaison, while the liaison also maintains its new position for the next round of position update.

9. The indoor toxic and harmful gas multi-potential source positioning system according to claim 1, characterized in that: The spatial positioning unit includes a ranging module and a modeling module, wherein the ranging module is used to obtain the spatial dimensions of the indoor environment and obtain the coordinates of its own position during the traceability detection process; the modeling module is used to construct the environmental space coordinates based on the distance information of the ranging unit, determine the current position of the detection instrument, and take a real-life picture and upload it as a model base map to facilitate the combination of the spatial modeling model and the coordinates of the actual scene.

10. A method for locating multiple potential sources of indoor toxic and harmful gases, characterized in that: The method adopts the indoor toxic and harmful gas multi-potential source positioning system according to any one of claims 1 to 9, comprising the following steps: Use spatial positioning units to perform spatial modeling for indoor environments; Use the point traceability unit to trace and update the measurement points in the modeling space; Use the spatial positioning unit to receive the updated position sent by the point traceability unit and guide the measurement point to move to the updated position; The detection unit is used to perform high-precision in-situ gas detection at the measurement point, and the detected indoor toxic and harmful gas concentration is fed back to the point traceability unit for subsequent location updates.

Images