Active and passive fusion traceability method for drainage basin pollutant discharge

Abstract

The invention discloses an active and passive fusion traceability method for drainage basin pollutant discharge, which creatively combines a passive traceability algorithm and an active traceability algorithm, broadens the channel of single traceability to obtain information, and in a pollution source tracking stage, an unmanned ship calls the passive traceability algorithm to invert to obtain pollution source position information, so that the pollution source position information can be accurately traced. And providing a motion direction for the active traceability as a global search strategy, feeding back monitoring information to the passive traceability after the unmanned ship arrives at a new position, updating source item information again by the passive traceability, and continuously performing information interaction between the active traceability and the passive traceability to realize pollution source tracking. And when the condition of falling into local optimum or deviating from the pollution zone occurs, the unmanned ship continues to call a passive traceability algorithm to provide a motion direction and step length for active traceability so as to achieve the purpose of jumping out of the local optimum or returning to the pollution zone, and the unmanned ship continues to carry out the pollution source tracking process until the unmanned ship successfully finds the position of the pollution source.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of water pollution tracing, and relates to a passive-active fusion tracing method for basin pollutant discharge. BACKGROUND

[0002] The emergency disposal of abnormal water environment in a basin is time-critical, and the scheme of checking the river discharge outlets one by one often needs to spend a lot of time, which is not feasible in practice. The primary task after pollution occurs is to determine the pollution source at the first time to restore the historical migration process of the pollutant and quickly predict the spread and diffusion of the pollutant. The multi-scale and multi-scene pollutant tracing technology and equipment of a basin can play a key role in conventional pollution control and sudden water pollution accident treatment, can efficiently track and locate the pollution source, and improve the polluted water quality and the water environment. At present, the research on sudden water pollution tracing at home and abroad is still in the exploratory and development stage, and with the perfection and maturity of environmental hydrology theory, the pollutant tracing theory is also gradually mature. Compared with the problem of predicting the concentration distribution of pollutants in a river channel according to known pollution sources and water quality parameters, the pollutant tracing is a kind of inverse problem in environmental hydrology, which is nonlinear and ill-posed.

[0003] At present, scholars at home and abroad have made a lot of research results in water pollution tracing. These research results mainly fall into two types of tracing methods, namely passive tracing algorithm and active tracing algorithm.

[0004] The passive tracing method uses the process observation value of the fixed monitoring station to inverse the pollution discharge position, discharge intensity and discharge time according to the migration and transformation law of the pollutant in the river channel. The core idea is to find the minimum difference between the actual water quality monitoring data and the output data of the mathematical model, so as to inversely deduce the discharge position, discharge amount and discharge time of the pollutant. This process can be regarded as the solution of the optimal value problem. For example, Zhu Song (Journal of Jiangsu University: Natural Science Edition, 2007, 28(3):4.) proposed a Bayesian algorithm for parameter identification of a water environment system based on a one-dimensional river water quality equation; Jiang Jiping (Frontiers of Environmental Science&Engineering, 2018, 12(5):1-16) carried out sensitivity analysis and inversion on the uncertainty parameters such as diffusion coefficient and flow rate by directly sampling, using regional sensitivity analysis method, identifiable figure and perturbation method; Wang Jiabiao and Lei Xiaohui (Journal of Hydraulic Engineering, 2015, 46(11):1280-1289.) constructed an optimization model with pollution source coordinates and discharge time as parameters based on water dynamic calculation and combining forward concentration probability density with reverse concentration probability density, and solved the model by using differential evolution algorithm. The active tracing method focuses on actively searching for the contaminated area. The basic principle is to use mobile devices such as unmanned ships carrying sensors to dynamically obtain target pollutant information, and to track the place with the highest pollution concentration by using efficient algorithms to achieve the purpose of tracing. This method can help locate the possible position of the pollution source, but usually requires more calculation and resource support. For example, Russell (Proceedings of the Australian Conference on Robotics and Automation[C]. 2003: 1-6.) was inspired by ants searching for food and proposed a pollution source positioning algorithm. An ant robot equipped with a pollutant concentration sensor was designed to search for the pollution source through "Z" type movement; Farrel (Adaptive Behavior, 2001, 9(3-4): 143-170.) proposed a water pollution tracing method based on behavior planning for chemical plume source pollution positioning problem. The entire search process can be divided into six behaviors, respectively: starting, finding pollution belt, tracking pollution belt, deviating, finding pollution belt again, and confirming the pollution source.

[0005] The passive tracing algorithm and the active tracing algorithm have the following problems: the passive tracing algorithm is mostly limited by uneven distribution of monitoring stations or the fact that there is not necessarily a monitoring station near the accident pollution source or surrounded by monitoring stations. Although the environmental monitoring quality can be improved by increasing the number of monitoring stations and observation frequency, there are still problems such as limited monitoring range, high cost, poor mobility, and the inversion accuracy depends on the model accuracy in practical application; the active tracing algorithm does not fully utilize the information of the monitoring stations, resulting in that when the pollutant concentration in the downstream of the river is too low, the sensors on both sides of the unmanned ship cannot monitor the significant difference in the concentration of the target pollutant, and secondly, the unmanned ship is easy to fall into local optimization because the search step is too small, and it is easy to deviate from the pollution belt because the search step is too large.

[0006] The present application combines the passive tracing algorithm and the active tracing algorithm, fully utilizes the monitoring information of the fixed monitoring stations and the unmanned ship, avoids the problem of not being able to find the position of the pollution source due to the limited model accuracy and the limited number of fixed monitoring stations in the passive tracing algorithm, and also solves the problem of being easy to fall into local optimization and deviate from the pollution belt in the active tracing algorithm. SUMMARY

[0007] The purpose of the present application is to make up for the technical deficiencies in the above-mentioned water pollution source tracing field, creatively combine active source tracing algorithm and passive source tracing algorithm, and propose a kind of active and passive fusion source tracing method for basin pollutant discharge.When the sensors on both sides of the unmanned ship cannot detect obvious differences, the passive source tracing algorithm is called to trace the source of water pollution according to the signals of the fixed monitoring station and the information of the unmanned ship.When the sensors on both sides of the unmanned ship detect obvious differences, the active source tracing algorithm is called to trace the source of pollution according to the concentration of pollutants monitored by the sensors on both sides of the unmanned ship.In the process of active source tracing, when the local optimum is reached or the pollution zone is left, the passive source tracing algorithm is called again to jump out of the local optimum and return to the pollution zone.

[0008] A kind of active and passive fusion source tracing method for basin pollutant discharge, comprising the following steps:

[0009] Step 1: set the positions of the fixed monitoring stations A, B and C, the threshold θ1, the threshold θ2 and the maximum iteration number genmax;

[0010] Step 2: the fixed monitoring station monitors the abnormal concentration of the target pollutant, and starts the source tracing program of the unmanned ship.The target pollutant sensors are symmetrically installed on the left and right sides of the unmanned ship, and the distance between the target pollutant sensors and the center of mass of the unmanned ship is L.

[0011] Step 3: record the change of the target pollutant concentration monitored by the fixed monitoring station, and call the passive source tracing algorithm to calculate the position of the pollution source The unmanned ship moves a step step towards the position;

[0012] Step 4: determine whether the target pollutant concentration detected by the sensors on both sides of the unmanned ship changes, if the target pollutant concentration is abnormal, go to step 5;Otherwise, return to step 3;

[0013] Step 5: combine the position information and the monitored target pollutant concentration information of the monitoring station and the unmanned ship, call the passive source tracing algorithm, and optimize to obtain a new pollution source position The unmanned ship moves a step step towards the position;

[0014] Step 6: determine whether the difference between the concentrations of pollutants monitored by the sensors on both sides of the unmanned ship is greater than the threshold θ1, if the difference between the concentrations of pollutants monitored by the sensors on both sides of the unmanned ship is greater than the threshold θ1, go to step 7;Otherwise, return to step 5;

[0015] Step 7: the unmanned ship calls the active source tracing algorithm according to the target pollutant concentration monitored by the concentration sensors on both sides, outputs the next position coordinate of the unmanned ship and goes to the position;

[0016] Step 8: Determine whether the unmanned ship is trapped in a local optimum. If the unmanned ship repeatedly wanders around a location, it is determined that the unmanned ship is trapped in a local optimum, and then return to step 5; otherwise, go to step 9;

[0017] Step 9: Determine whether the unmanned ship has left the pollution zone. If the unmanned ship has left the pollution zone, return to step 5; otherwise, go to step 10;

[0018] Step 10: Determine whether the target pollutant concentration value measured by the sensor on the unmanned ship is greater than the threshold value θ2. If it is greater than the threshold value θ2, it is determined that the source of the pollutant has been found; if it is less than the threshold value θ2, go to step 11;

[0019] Step 11: Determine whether the maximum number of iterations genmax has been reached. If the maximum number of iterations is reached, it is determined that the source of the pollutant cannot be found; otherwise, return to step 7.

[0020] The active tracing algorithm mentioned in step 7 above uses the tentacle search algorithm, and its iteration process steps are as follows:

[0021] Step 1: Randomly rotate any angle with the center of mass of the unmanned ship, and the unit vector of the direction in which the sensor is pointing at this time is:

[0022]

[0023] Where rands(2,1) represents a random value of [0, 1] selected in the direction of the river flow and perpendicular to the direction of the river flow, which is used to calculate the direction unit vector;

[0024] The positions of the left and right sensors of the unmanned ship are calculated as:

[0025]

[0026] Where x is the position of the current center of mass of the unmanned ship, x L is the position of the left sensor of the unmanned ship, and x R is the position of the right sensor of the unmanned ship;

[0027] Step 2: Obtain the left concentration value C L and the right concentration value C R , compare the concentration values on both sides, and take the higher value;

[0028] Step 3: If the left concentration value is high, the next position is output as:

[0029]

[0030] If the right concentration value is high, the next position is output as:

[0031]

[0032] Where δ is the step size, and x t x is the current position of the unmanned vessel. t+1 To determine the next location of the unmanned vessel;

[0033] The passive source tracing algorithm mentioned in steps 3 and 5 above uses the target pollutant concentration and location information monitored by fixed monitoring stations and unmanned vessels to invert the pollution emission location, emission intensity and emission time. By optimizing the objective function, the location, emission intensity and emission time of the water pollution accident pollution source that is closest to the calculated value and the measured value are obtained. This optimization search process is completed by a genetic algorithm, and the steps are as follows;

[0034] Step 1: Determine the location range, emission intensity range, and emission time range of the pollution source;

[0035] Step 2: Given the population size N and boundary constraints, generate the first generation population:

[0036] Step 3: Substitute the population information into the two-dimensional diffusion model to obtain the calculated and predicted concentration value, and substitute it with the actual monitored pollutant concentration into the optimization objective function to obtain the fitness value. Based on the fitness value, determine whether the stopping condition is met. If so, stop and output the optimal solution; otherwise, continue the operation.

[0037] The objective function to be optimized is:

[0038]

[0039] n represents the number of monitoring times, and m represents the number of monitoring locations. This represents the theoretical pollutant concentration at position j at time i. This represents the actual pollutant concentration value monitored by the fixed monitoring station at position j at time i. This represents the actual pollutant concentration value monitored by the unmanned vessel at position j at time i.

[0040] Step 4: Perform selection, crossover, and mutation operations on the population to obtain the next generation population, and return to step 3;

[0041] The beneficial effects of this invention are as follows:

[0042] This invention creatively combines passive and active source tracing algorithms, broadening the channels for obtaining information through single source tracing. The entire search process can be decomposed into six behaviors: initiation, entering the pollution zone, tracking the pollution source, getting trapped in a local optimum or deviating from the pollution zone, returning to the pollution zone, and confirming the pollution source. In the initial stage, due to the unclear abnormal concentration of target pollutants in the downstream area, the unmanned surface vessel (USV) uses the pollution source location information obtained from the passive source tracing algorithm as a global search strategy to provide the direction and step size for active source tracing until the USV enters the pollution zone. During the pollution source tracking stage, the USV continues to use the pollution source location information obtained from the passive source tracing algorithm to provide the direction and step size for active source tracing. After reaching a new location, the USV feeds back the monitoring information to the passive source tracing, which then updates the source term information again. Active and passive source tracing continuously exchange information to achieve pollution source tracking. When the USV gets stuck in a local optimum or deviates from the pollution zone, it uses the pollution source location information obtained from the passive source tracing algorithm to provide the direction and step size for active source tracing. This allows it to escape the local optimum or return to the pollution zone, and then continues the pollution source tracking process until the USV successfully finds the pollution source location. Attached Figure Description

[0043] Figure 1 A flowchart of a combined active and passive source tracing method for watershed pollutant emissions.

[0044] Figure 2 This is an iterative simulation diagram of an active water pollution source tracing method based on the longhorn beetle whisker search algorithm.

[0045] Figure 3 This is an iterative simulation diagram of a water pollution source tracing method that combines active and passive approaches based on genetic algorithms and longhorn beetle whisker search algorithms. Detailed Implementation

[0046] The present invention will be further described below with reference to the accompanying drawings and specific preferred embodiments, but this does not limit the scope of protection of the present invention.

[0047] A combined active and passive source tracing method for watershed pollutant emissions, the process of which is as follows: Figure 1 As shown, it includes the following steps:

[0048] Step 1: Set the fixed monitoring station location, threshold θ1, threshold θ2, maximum number of iterations genmax, and construct the concentration field;

[0049] In this embodiment, the locations of the three fixed monitoring stations are (190, 50), (300, 45), and (380, 30), respectively. The threshold θ1 is 0.02, the threshold θ2 is 14.5, the maximum number of iterations genmax is 100, the pollutant emission rate in the constructed concentration field is 100 g / s, and the lateral diffusion parameter is 0.5 m.2 / s, longitudinal diffusion parameter is 0.2m 2 / s, the lateral velocity is 0.5m / s, the longitudinal velocity is 0.015m / s, the pollution source location is (0,0), and the monitoring time is 1200s;

[0050] Step 2: When the fixed monitoring station detects an abnormal concentration of the target pollutant, it initiates the source tracing procedure of the unmanned vessel. Target pollutant sensors are symmetrically installed on the left and right sides of the unmanned vessel, and the distance between the pollutant sensors and the center of mass of the unmanned vessel is L.

[0051] In this embodiment, the distance between the pollutant sensor and the center of mass of the unmanned vessel is 2.

[0052] Step 3: Record the changes in the concentration of the target pollutant monitored by the fixed monitoring station, and use the passive source tracing algorithm to calculate the location of the pollution source. The unmanned boat moves one step toward that location;

[0053] In this embodiment, the passive source tracing algorithm calculates the location of the pollution source for the first time. The coordinates are (84.446, 19.896), the initial position of the unmanned vessel is (400, 25), and the step size is 10.

[0054] Step 4: Determine whether the target pollutant sensors on both sides of the unmanned vessel detect abnormal pollutant concentrations. If a change in target pollutant concentration is detected, proceed to Step 5; otherwise, return to Step 3.

[0055] Step 5: Combining the location information of the monitoring station and the unmanned vessel with the monitored target pollutant concentration information, the passive source tracing algorithm is invoked to optimize and obtain the location of the new pollution source. The unmanned boat moves one step toward that location;

[0056] Step 6: Determine whether the concentration difference of the target pollutant detected by the sensors on both sides of the unmanned vessel is greater than the threshold θ1. If the detected concentration difference of the target pollutant is greater than the threshold θ1, proceed to step 7; otherwise, return to step 5.

[0057] Step 7: The unmanned surface vessel (USV) detects the concentration of the target pollutant based on the concentration sensors on both sides, calls the active source tracing algorithm, outputs the coordinates of the USV's next location, and proceeds there.

[0058] Step 8: Determine if the unmanned vessel is trapped in a local optimum. If the unmanned vessel repeatedly hovers around a position, it is determined that the unmanned vessel is trapped in a local optimum, and then return to step 5; otherwise, proceed to step 9.

[0059] Step 9: Determine whether the unmanned vessel has left the pollution zone. If the unmanned vessel has left the pollution zone, return to step 5; otherwise, proceed to step 10.

[0060] Step 10: Determine whether the concentration of the target pollutant measured by the sensors on the unmanned vessel is greater than the threshold θ2. If it is greater than the threshold θ2, then the source of the pollutant has been found; if it is less than the threshold θ2, proceed to step 11.

[0061] Step 11: Determine if the maximum number of iterations genmax has been reached. If the maximum number of iterations has been reached, it is determined that the source of the contaminant cannot be found; otherwise, return to step 7.

[0062] The active source tracing algorithm mentioned in step 7 above uses the longhorn beetle whisker search algorithm, and its iterative process steps are as follows:

[0063] Step 1: Rotate the unmanned vessel randomly by any angle, with its center of mass as the center. The unit vector pointing towards the sensor at this point is:

[0064]

[0065] rands(2,1) represents a random selection of a value [0,1] in the direction of river flow and the direction perpendicular to the direction of river flow, used to calculate the direction unit vector;

[0066] The calculated positions of the left and right sensors of the unmanned vessel are as follows:

[0067]

[0068] Where x is the current position of the unmanned vessel's center of mass, x L x represents the position of the sensor on the left side of the unmanned vessel. R This indicates the location of the sensor on the right side of the unmanned vessel.

[0069] Step 2: Obtain the concentration value C on the left side of the unmanned vessel. L and the concentration value C on the right. R Compare the concentration values ​​on both sides and take the higher value;

[0070] Step 3: If the concentration value on the left is high, then output the next position as:

[0071]

[0072] If the concentration value on the right is high, then the next position will be output as:

[0073]

[0074] In this embodiment, the step size δ is 10, where x t x is the current position of the unmanned vessel. t+1 To determine the next location of the unmanned vessel;

[0075] The passive source tracing algorithm mentioned in steps 3 and 5 above uses the target pollutant concentration and location information monitored by fixed monitoring stations and unmanned vessels to invert the pollution emission location, emission intensity, and emission time. By optimizing the objective function, it obtains the location, emission intensity, and emission time of the water pollution source that best matches the measured values. This optimization search process is completed using a genetic algorithm, and the steps are as follows:

[0076] Step 1: Determine the location range, emission intensity range, and emission time range of the pollution source;

[0077] In this embodiment, the x-axis range of the pollution source location is [-50, 200], the y-axis range of the pollution source location is [-20, 20], the pollution intensity range is [0, 100], and the emission time range is [0, 1500].

[0078] Step 2: Given the population size N and boundary constraints, generate the first generation population;

[0079] In this embodiment, the population size N is 800;

[0080] Step 3: Substitute the population information into the two-dimensional diffusion model to obtain the calculated and predicted concentration value, and substitute it with the actual monitored pollutant concentration into the optimization objective function to obtain the fitness value. Based on the fitness value, determine whether the stopping condition is met. If so, stop and output the optimal solution; otherwise, continue the operation.

[0081] The two-dimensional diffusion model in this embodiment:

[0082]

[0083] The objective function to be optimized is:

[0084]

[0085] Step 4: Perform selection, crossover, and mutation operations on the population to obtain the next generation population, and return to step 3;

[0086] In this embodiment, the performance of the combined active-passive algorithm for pollution source localization is analyzed by comparing it with the beetle whisker search algorithm and the genetic algorithm. Nine sets of simulation backgrounds with different environmental factors were set, including pollution source emission rates of 100 g / s, 200 g / s, and 300 g / s, and river flow velocities of 0.5 m / s, 1 m / s, and 1.5 m / s. In each set of experiments, the beetle whisker search algorithm, the genetic algorithm, and the combined active-passive algorithm were run 100 times each. If the pollution source was found within a radius of 15 m, the experiment was considered successful. The simulation results are shown in Table 1. Table 1. Simulation results of three source tracing methods under different conditions

[0087] The results show that the combined active and passive algorithm outperforms the beetle whisker search algorithm and the genetic algorithm in overall performance. Not only does it achieve a success rate of over 92%, but its average number of iterations is also much smaller than that of the beetle whisker search algorithm.

[0088] like Figure 2 As shown, the active source tracing method based on the longhorn beetle whisker search algorithm gets stuck in a local optimum during the source tracing process, resulting in the inability to find the source of pollutants. Figure 3 The water pollution source tracing method shown combines passive source tracing based on genetic algorithm and active source tracing based on beetle whisker search algorithm. It makes full use of monitoring data from fixed monitoring stations and unmanned vessels, calls the passive source tracing algorithm to obtain source term information, solves the problem of getting stuck in local optima during active source tracing, and thus achieves the goal of accelerating global optimization.

[0089] The above description is only the most effective embodiment of the present invention. It should be noted that for those skilled in the art, appropriate improvements and modifications can be made without departing from the working principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A combined active and passive source tracing method for watershed pollutant emissions, characterized in that, Includes the following steps: Step 1: Set fixed monitoring station locations A, B, and C, threshold θ1, threshold θ2, and maximum number of iterations genmax; Step 2: When the fixed monitoring station detects an abnormal concentration of the target pollutant, it initiates the source tracing procedure of the unmanned vessel. Target pollutant sensors are symmetrically installed on the left and right sides of the unmanned vessel, and the distance between the target pollutant sensors and the center of mass of the unmanned vessel is L. Step 3: Record the changes in the concentration of the target pollutant monitored by the fixed monitoring station, and use the passive source tracing algorithm to calculate the location of the pollution source. The unmanned boat moves one step toward that location; Step 4: Determine whether the target pollutant sensors on both sides of the unmanned vessel have detected changes in the concentration of the target pollutant. If an abnormal concentration of the target pollutant is detected, proceed to Step 5; otherwise, return to Step 3. Step 5: Combining the location information of the monitoring station and the unmanned vessel with the monitored target pollutant concentration information, the passive source tracing algorithm is invoked to optimize and obtain the new pollution source location. The unmanned boat moves one step toward that location; Step 6: Determine whether the concentration difference of the target pollutant detected by the sensors on both sides of the unmanned vessel is greater than the threshold θ1. If the detected concentration difference of the target pollutant is greater than the threshold θ1, proceed to step 7; otherwise, return to step 5. Step 7: The unmanned surface vessel (USV) detects the concentration of the target pollutant based on the concentration sensors on both sides, calls the active source tracing algorithm, outputs the coordinates of the USV's next location, and proceeds there. Step 8: Determine if the unmanned vessel is trapped in a local optimum. If the unmanned vessel repeatedly hovers around a position, it is determined that the unmanned vessel is trapped in a local optimum, and then return to step 5; otherwise, proceed to step 9. Step 9: Determine whether the unmanned vessel has left the pollution zone. If the unmanned vessel has left the pollution zone, return to step 5; otherwise, proceed to step 10. Step 10: Determine whether the concentration of the target pollutant measured by the sensors on the unmanned vessel is greater than the threshold θ2. If it is greater than the threshold θ2, then the source of the pollutant has been found; if it is less than the threshold θ2, proceed to step 11. Step 11: Determine if the maximum number of iterations genmax has been reached. If the maximum number of iterations has been reached, it is determined that the source of the contaminant cannot be found; otherwise, return to step 7.

2. The active and passive source tracing method for watershed pollutant emissions according to claim 1, characterized in that... The active source tracing algorithm mentioned in step 7 uses the longhorn beetle whisker search algorithm, and its iterative process steps are as follows: Step 1: Rotate the unmanned vessel randomly by any angle, with its center of mass as the center. The unit vector pointing towards the sensor at this point is: rands(2,1) represents a random selection of a value [0,1] in the direction of river flow and the direction perpendicular to the direction of river flow, used to calculate the direction unit vector; The calculated positions of the left and right sensors of the unmanned vessel are as follows: Where x is the current position of the unmanned vessel's center of mass, x L x represents the position of the sensor on the left side of the unmanned vessel. R This indicates the location of the sensor on the right side of the unmanned vessel. Step 2: Obtain the concentration value C on the left side of the unmanned vessel. L and the concentration value C on the right. R Compare the concentration values ​​on both sides and take the higher value; Step 3: If the concentration value on the left is high, then output the next position as: If the concentration value on the right is high, then the next position will be output as: Where δ is the step size, and x t x is the current position of the unmanned vessel. t+1 This indicates the next location for the unmanned vessel.

3. The active and passive source tracing method for watershed pollutant emissions according to claim 1, characterized in that... The passive source tracing algorithm mentioned in steps 3 and 5 uses the target pollutant concentration and location information monitored by fixed monitoring stations and unmanned vessels to invert the pollution emission location, emission intensity, and emission time. By optimizing the objective function, the location of the water pollution source that is closest to the calculated value and the measured value can be obtained. This optimization search process is completed by a genetic algorithm, and the steps are as follows: Step 1: Determine the location range, emission intensity range, and emission time range of the pollution source; Step 2: Given the population size N and boundary constraints, generate the first generation population; Step 3: Substitute the population information into the two-dimensional diffusion model to obtain the calculated and predicted concentration value, and substitute it with the actual monitored pollutant concentration into the optimization objective function to obtain the fitness value. Based on the fitness value, determine whether the stopping condition is met. If it is met, stop and output the optimal solution; otherwise, continue the operation. The objective function to be optimized is: n represents the number of monitoring times, and m represents the number of monitoring locations. This represents the theoretical pollutant concentration at position j at time i. This represents the actual pollutant concentration value monitored by the fixed monitoring station at position j at time i. This represents the actual pollutant concentration value monitored by the unmanned surface vessel at position j at time i; Step 4: Perform selection, crossover, and mutation operations on the population to obtain the next generation population, and return to step 3.

Images