Unmanned ship water pollution tracing method and system

Abstract

The present application relates to the technical field of water pollution tracing, and particularly relates to a method and system for tracing water pollution by an unmanned ship, which comprises the following steps: constructing an improved dynamic behavior tree decision model after starting a tracing task, which comprises three parallel behavior sub-trees of tracing, obstacle avoidance and returning; collecting the concentrations of pollutants, the distances of obstacles and the remaining power of the unmanned ship in real time in four directions of front, back, left and right; calculating the behavior weights and normalizing them in each sampling period; judging based on behavior switching logic, which introduces hysteresis constraint and time constraint, and requires that the normalized weight of the behavior to be switched is higher than the sum of the current behavior weight and a threshold increment, and the duration of this state exceeds the anti-shake time, and if the conditions are met, the behavior with the highest weight is switched to; finally, the corresponding algorithm is called to execute the action; the system comprises the modules of main control, concentration detection, obstacle avoidance, positioning, driving, power, wireless communication and host computer control, which is beneficial to solve the problems of inflexible task switching and easy decision shock caused by instantaneous interference or close weights of the traditional behavior tree in complex water areas, and improves the autonomy, task execution efficiency and overall system stability of the unmanned ship tracing.

Description

Technical Field

[0002] This invention belongs to the field of water pollution source tracing technology, specifically relating to an unmanned vessel water pollution source tracing method and system. Background Technology

[0004] As my country's industrialization continues, illegal discharge of pollutants by enterprises persists despite repeated bans. Some companies directly discharge untreated industrial wastewater into rivers through sewage outlets, causing numerous water pollution incidents and posing a serious threat to the ecological environment of river basins and public health. Currently, water pollution monitoring mainly relies on fixed monitoring stations, but this method has significant limitations: the coverage of monitoring stations is limited, and if a pollution incident occurs in an area without monitoring stations, it is often difficult to detect in a timely manner; the determination of the pollution source location relies heavily on mathematical models, the accuracy of which is affected by parameter settings and actual hydrological conditions, and the calculated results are prone to deviating from the actual discharge location; in addition, fixed monitoring stations have fixed locations and lack flexibility, making it impossible to actively track pollution sources.

[0005] Compared to fixed monitoring methods, unmanned surface vessels (USVs) have been gradually applied to water pollution source tracing tasks due to their advantages of maneuverability, lower cost, and convenient deployment. However, existing USV-based source tracing methods still have shortcomings when dealing with complex river environments: most systems use fixed task priorities, which lack flexibility when encountering complex situations; when the weight values ​​of two actions are close, the system lacks a hysteresis control mechanism, causing the USV to frequently switch between different actions due to small fluctuations in sensor data, resulting in decision oscillations; for transient obstacles or environmental noise accidentally detected by ultrasonic sensors, the system lacks a time-constrained delay judgment mechanism, which may trigger unnecessary obstacle avoidance actions, interrupt the source tracing path, and reduce source tracing efficiency.

[0006] In the prior art, the invention with patent number CN202510674717.X discloses a water turbine generator bore inspection robot and its operation method based on dynamic behavior tree. It is mainly suitable for enclosed spaces and has insufficient environmental adaptability to open environments. It does not solve the anti-shaking problem of dynamic competition of multiple tasks in open waters. The invention with patent number CN202510381888.6 discloses a water pollution source tracing analysis method based on intelligent unmanned boat. Its decision logic relies on fixed process control. When facing complex waters, it cannot handle the real-time priority conflict between obstacle avoidance and source tracing tasks, and it does not involve a mechanism to prevent decision oscillation. Summary of the Invention

[0008] The purpose of this invention is to provide a method and system for tracing water pollution sources using unmanned vessels, which is beneficial to improving the autonomous decision-making ability and execution efficiency of unmanned vessels in pollution source tracing tasks.

[0009] To achieve the above objectives, the technical solution of the present invention is as follows:

[0010] A method for tracing the source of water pollution from unmanned surface vessels, characterized by comprising the following steps:

[0011] S1. Initiate the pollution source tracing task;

[0012] S2. Construct an improved dynamic behavior tree decision model, including three parallel behavior subtrees: source tracing, obstacle avoidance, and return to home.

[0013] S3. Real-time collection of pollutant concentration, obstacle distance, and remaining battery power in four directions (front, rear, left, and right) of the unmanned vessel;

[0014] S4. Based on the environmental data collected in S3, the unmanned vessel calculates the source tracing weight, obstacle avoidance weight, and return weight in each sampling period T based on the improved dynamic behavior tree decision model, and performs normalization processing.

[0015] S5. Based on the behavior switching logic of the improved dynamic behavior tree decision model, a judgment is made. If the behavior switching logic is satisfied, the behavior subtree with the highest weight is executed; otherwise, the current behavior continues to be executed.

[0016] S6. Based on the final decision in S5, the unmanned vessel executes the control algorithm corresponding to the behavior subtree to complete the corresponding action.

[0017] Furthermore, the improved dynamic behavior tree decision model includes: a root node, behavior subtree nodes, and action nodes. The root node calculates the normalized weight values ​​of the source tracing, obstacle avoidance, and return navigation behavior subtrees based on the environmental data obtained by the unmanned vessel's main control module. When the normalized weight value of a non-current behavior satisfies the behavior switching logic, a behavior switch is triggered, thereby activating a behavior subtree in the next layer. The behavior subtree nodes include parallel source tracing behavior subtrees, obstacle avoidance behavior subtrees, and return navigation behavior subtrees. The entry point of each subtree is a sequential node, used to trigger the action nodes under that subtree in sequence. The action node is the terminal node, used to execute the control algorithm corresponding to the action.

[0018] Furthermore, the behavior switching logic includes the following two constraints:

[0019] S51. Hysteresis constraint requires that the normalized weight value of the behavior to be switched must be continuously higher than the sum of the normalized weight value of the currently executed behavior and the set threshold increment Δ;

[0020] S52. The time constraint requires that, under the premise of satisfying the hysteresis constraint, the duration of this weighting relationship exceeds the set anti-shake time T. switch .

[0021] This invention provides an unmanned vessel water pollution source tracing system, comprising:

[0022] The unmanned surface vessel (USV) body, and the main control module, concentration detection module, obstacle avoidance detection module, power monitoring module, positioning module, drive module and wireless communication module installed on the USV body; it also includes a host computer control module that communicates with the wireless communication module.

[0023] The unmanned surface vessel (USV) is equipped with several modules, including: a concentration detection module for acquiring pollutant concentration data in the front, rear, left, and right directions; an obstacle avoidance detection module for acquiring obstacle distance data in the front, rear, left, and right directions; a power monitoring module for acquiring remaining power data; and a positioning module for acquiring the USV's location information. All of these modules are connected to the main control module.

[0024] The main control module stores and runs an improved dynamic behavior tree decision model, which includes three parallel behavior subtrees: source tracing, obstacle avoidance, and return. Within each sampling period T, the main control module calculates the source tracing weight, obstacle avoidance weight, and return weight based on pollutant concentration data, obstacle distance data, and remaining power data, and performs normalization processing. Based on the preset behavior switching logic, it determines the behavior subtree to be executed and then outputs control commands to the drive module to drive the unmanned vessel to perform the corresponding actions.

[0025] The host computer control module is used for task initialization, remote monitoring, environmental data visualization and trajectory display, while the navigation behavior decision-making and control of the unmanned vessel are completed by the main control module.

[0026] Specifically, the main control module uses an embedded microcontroller as the core control unit, and its memory size is sufficient to store and run the behavior control program of the unmanned vessel. It is used to calculate the weight value of each behavior within the sampling period and output control instructions and execute them according to the behavior switching logic.

[0027] Specifically, the concentration detection module consists of four water quality sensors, which are fixed at the bow, stern, left side and right side of the ship, respectively. The detection end is submerged about 10 cm below the water surface and is used to detect water quality parameters in the water in real time.

[0028] Specifically, the obstacle avoidance module consists of four ultrasonic sensors, installed at a height of about 30cm above the water surface, with a detection range of 2cm to 400cm. When an obstacle less than 20cm away is detected, an obstacle avoidance signal is sent to the main control module to detect the distance of obstacles around the unmanned vessel in real time.

[0029] Specifically, the power monitoring module is used to monitor the remaining power of the unmanned ship's onboard battery in real time;

[0030] Specifically, the positioning module is implemented by an RTK module, which has high positioning accuracy and can achieve centimeter-level positioning, and is used to locate the position of the unmanned vessel in real time;

[0031] Specifically, the drive module is used to implement the control commands issued by the unmanned vessel's main control module;

[0032] Specifically, the wireless communication module is used for real-time communication between the unmanned vessel's main control module and the host computer control module;

[0033] Furthermore, the concentration detection module, the obstacle avoidance module, the positioning module, the drive module, and the power module are respectively mounted on the main control module;

[0034] This invention proposes an unmanned surface vessel (USV) water pollution source tracing method and system. It simultaneously incorporates hysteresis and time constraints into the behavior switching logic of a dynamic behavior tree. Unlike existing behavior tree schemes that rely on immediate condition judgments or fixed priorities, the hysteresis constraint effectively suppresses frequent behavior switching caused by sensor data fluctuations or similar behavior weight values, improving system stability and anti-interference capabilities. The time constraint ensures behavior switching is based on continuous and effective environmental changes, avoiding the impact of instantaneous environmental interference on decision-making results. The constructed improved dynamic behavior tree model places the three tasks of source tracing, obstacle avoidance, and return to shore within a competitive decision-making framework, enabling the USV to make autonomous, flexible, and intelligent decisions in complex waters, thereby improving the efficiency of source tracing tasks. Deeply integrating the improved dynamic behavior tree model with the USV source tracing system enhances the overall intelligent decision-making capability and engineering applicability of the system, facilitating the practical deployment and widespread application of the USV water pollution source tracing method. Attached Figure Description

[0036] Figure 1 This is a flowchart of the steps of the unmanned vessel water pollution source tracing method in this embodiment of the invention;

[0037] Figure 2 This is a structural diagram of the improved dynamic behavior tree decision model in this embodiment of the invention;

[0038] Figure 3 This is a schematic diagram of the unmanned vessel water pollution source tracing system in an embodiment of the present invention; Detailed Implementation

[0040] Various embodiments of the present invention will now be clearly and completely described with reference to the accompanying drawings. The embodiments described with reference to the drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0041] To fully illustrate the technical solution of the present invention, combined with Figure 1 The illustrated process provides a preferred implementation method that combines specific parameters with the operational flow:

[0042] In this embodiment, the key parameters of the improved dynamic behavior tree decision model are set as follows: the weight coefficients were determined through multiple water area tests and adjustments, and the source tracing weight coefficient k... s = 2.5, obstacle avoidance weight coefficient k o = 8.0, return weighting coefficient k p = 3.0, behavior switching threshold increment Δ = 0.15, stabilization time T switch = 1.0 seconds, the system sampling period T = 0.2 seconds, the weighting coefficient k s k o k p and threshold Δ, stabilization time T switch The system sampling period T is determined based on different ship hull performance and mission water environment.

[0043] W s The calculation formula is: C max C represents the maximum pollutant concentration in the four directions: front, back, left, and right. avg This represents the average concentration of pollutants in four directions. W o The calculation formula is: D min W represents the minimum distance to obstacles in the four directions: front, back, left, and right. p The calculation formula is: Where P is the current battery level of the unmanned vessel, P max This represents the maximum battery capacity.

[0044] As one implementation method, see Figure 1 and Figure 2 As shown, a method for tracing the source of water pollution from unmanned vessels includes the following steps:

[0045] S1. Initiate the pollution source tracing task;

[0046] S2. Construct an improved dynamic behavior tree decision model, including three parallel behavior subtrees: source tracing, obstacle avoidance, and return to home.

[0047] S3. Real-time collection of pollutant concentration, obstacle distance, and remaining battery power in four directions (front, rear, left, and right) of the unmanned vessel;

[0048] S4. Based on the environmental data collected in S3, the unmanned vessel calculates the source tracing weight, obstacle avoidance weight, and return weight in each sampling period T based on the improved dynamic behavior tree decision model, and performs normalization processing.

[0049] S5. Based on the behavior switching logic of the improved dynamic behavior tree decision model, a judgment is made. If the behavior switching logic is satisfied, the behavior subtree with the highest weight is executed; otherwise, the current behavior continues to be executed.

[0050] S6. Based on the final decision in S5, the unmanned vessel executes the control algorithm corresponding to the behavior subtree to complete the corresponding action.

[0051] Specifically, in S1 of this embodiment, after the unmanned vessel main control module receives the source tracing task from the host computer, it starts from the starting point and the initial state is to execute the "source tracing behavior".

[0052] Specifically, in S3 of this embodiment, the pollutant concentrations in four directions (front, rear, left, and right) of the unmanned vessel are collected in real time. The highest concentration is currently detected in front. max =150 μS / cm, C avg =120 μS / cm; Real-time monitoring of obstacle distance data in four directions (front, rear, left, and right) of the unmanned vessel to obtain D min =300cm; Real-time data shows the unmanned vessel's remaining battery power is 98%, indicating sufficient power.

[0053] Specifically, in S4 of this embodiment, the source tracing behavior weight W is calculated within each sampling period of 0.2s. s Obstacle avoidance weight W o Return weight W p W s = 3.125, W o = 0.400, W p = 0.075, for W s W o W p Normalize the weights so that their sum equals 1. The normalized weights are W. s' = 0.85, W o' = 0.10, W p' = 0.05;

[0054] Specifically, in S5 of this embodiment, the behavior switching logic of the improved dynamic behavior tree decision model is used for judgment. The behavior with the highest weight is the tracing behavior, which is consistent with the currently executed behavior. Therefore, there is no need to trigger the behavior switching logic, and the unmanned vessel maintains the tracing behavior and continues to execute. When an obstacle appears in the navigation path, causing D to... min The W plummeted to 25cm. o W increases sharply after normalization o' = 0.70, at this point, the weight W of the tracing behavior is... s' = 0.25, obstacle avoidance weight Wo' Greater than the current source tracing weight W s' The sum of the increments Δ and the preset threshold is used to start timing, and the system continuously samples and calculates W over the next 1.0 second. o' And determine whether it remains above W. s' + Δ, if the condition lasts for more than 1.0 second, the improved dynamic behavior tree decision model triggers behavior switching;

[0055] Specifically, in S6 of this embodiment, the unmanned vessel switches from "source tracing behavior" to "obstacle avoidance behavior" and calls the obstacle avoidance algorithm to plan a detour path.

[0056] It is understood that, in this embodiment, the hysteresis constraint specifically refers to the obstacle avoidance weight W. o′ Higher than the current traceability weight W s′ The sum of Δ and the time constraint specifically means that the state satisfying the hysteresis constraint must be maintained continuously for more than 1.0 seconds before the system will trigger the behavior switching logic, switching from source tracing behavior to obstacle avoidance behavior.

[0057] Furthermore, as one implementation method, see [link to implementation details]. Figure 2 As shown, an improved dynamic behavior tree decision model adopts a three-layer network structure, including a root node 101, behavior subtree nodes 102 and action nodes 103. The first layer is the root node 101, the second layer is the behavior subtree nodes 102, and the third layer is the action nodes 103.

[0058] Specifically, the first layer is the root node 101, which calculates the normalized weight values ​​of the source tracing, obstacle avoidance and return navigation behavior subtrees in real time based on the environmental information obtained by the unmanned vessel main control module. When the normalized weight value of a certain non-current behavior satisfies the behavior switching logic, the behavior switching is triggered, thereby activating a behavior subtree in the next layer.

[0059] Specifically, the second layer consists of behavior subtree nodes 102, which include parallel source tracing behavior subtrees, obstacle avoidance behavior subtrees, and return navigation behavior subtrees. The entry point of each subtree is a sequence node, which is used to trigger the action nodes 103 under that subtree in sequence.

[0060] Specifically, the third layer is action node 103, which is the end node of the behavior tree and is used to execute the control algorithm corresponding to the action.

[0061] It is understood that this invention introduces an improved dynamic behavior tree model, which divides the behavior of unmanned vessels into three independent behavior subtrees: source tracing, obstacle avoidance, and return. This allows for dynamic adjustment of the weights of each behavior based on environmental information, enabling parallel evaluation and behavior switching. In this way, unmanned vessels can have stronger autonomous decision-making capabilities and behavioral flexibility in complex aquatic environments, thereby improving the overall execution efficiency and success rate of water pollution source tracing tasks.

[0062] As an implementation method, the behavior switching logic described in S5 includes the following two constraints:

[0063] S51. Hysteresis constraints require that the normalized weight value of the behavior to be switched must be higher than the sum of the normalized weight value of the currently executing behavior and the set threshold increment Δ;

[0064] S52. The time constraint requires that, under the premise of satisfying the hysteresis constraint, the duration of this weighting relationship exceeds the set anti-shake time T. switch ;

[0065] It is understood that the method of the present invention introduces behavior switching logic containing hysteresis constraints and time constraints into the improved dynamic behavior tree model, thereby avoiding frequent behavior switching by unmanned vessels due to fluctuations in environmental information. In this way, it is beneficial to improve the stability and continuity of the unmanned vessel in the behavior decision-making process, enhance the decision rationality and anti-interference ability of the system operation, and improve its practicality in complex aquatic environments.

[0066] As one implementation method, see Figure 3 As shown, an unmanned surface vessel (USV) water pollution tracing system is provided, including: an USV body, and a main control module 1, a concentration detection module 2, an obstacle avoidance detection module 3, a power monitoring module 4, a positioning module 5, a drive module 6, and a wireless communication module 7 installed on the USV body; it also includes a host computer control module 8 that is communicatively connected to the wireless communication module;

[0067] Among them, the concentration detection module 2 is used to acquire pollutant concentration data in the front, back, left and right directions of the unmanned vessel; the obstacle avoidance detection module 3 is used to acquire distance data of obstacles around the unmanned vessel; the power monitoring module 4 is used to acquire the remaining power data of the unmanned vessel; the positioning module 5 is used to acquire the location information of the unmanned vessel; all of the above modules are connected to the main control module.

[0068] The main control module 1 stores and runs an improved dynamic behavior tree decision model, which includes three parallel behavior subtrees: source tracing, obstacle avoidance, and return. In each sampling period T, the main control module 1 calculates the source tracing weight, obstacle avoidance weight, and return weight based on pollutant concentration data, obstacle distance data, and remaining power data, and performs normalization processing. It then determines the behavior subtree to be executed according to the preset behavior switching logic and outputs control commands to the drive module to drive the unmanned vessel to perform the corresponding actions.

[0069] The host computer control module 8 is used for task initialization, remote monitoring, environmental data visualization and trajectory display, while the navigation behavior decision-making and control of the unmanned vessel are completed by the main control module.

[0070] Specifically, the main control module 1 uses an embedded microcontroller as the core control unit, and its memory size is sufficient to store and run the behavior control program of the unmanned vessel. It is used to calculate the weight value of each behavior within the sampling period and output control instructions and execute them according to the behavior switching logic.

[0071] Specifically, the concentration detection module 2 consists of four water quality sensors, which are fixed at the bow, stern, left side and right side of the ship respectively. The detection end is submerged about 10cm below the water surface and is used to detect water quality parameters in the water in real time.

[0072] Specifically, the obstacle avoidance module 3 consists of four ultrasonic sensors, installed at a height of about 30cm above the water surface, with a detection range of 2cm to 400cm. When an obstacle less than 20cm away is detected, it sends an obstacle avoidance signal to the main control module 1 to detect the distance of obstacles around the unmanned vessel in real time.

[0073] Specifically, the power monitoring module 4 is used to monitor the remaining power of the unmanned ship's onboard battery in real time;

[0074] Specifically, the positioning module 5 is implemented by an RTK module, which has high positioning accuracy and can achieve centimeter-level positioning, and is used to locate the position of the unmanned vessel in real time;

[0075] Specifically, the drive module 6 is used to implement the control commands issued by the unmanned vessel's main control module;

[0076] Specifically, the wireless communication module 7 is used for real-time communication between the unmanned vessel main control module 1 and the host computer control module 8;

[0077] Furthermore, the concentration detection module 2, the obstacle avoidance module 3, the positioning module 4, the drive module 5, and the power module 6 are respectively mounted on the main control module 1;

[0078] It is understandable that the unmanned vessel water pollution source tracing system, by combining multi-source environmental perception with an improved dynamic behavior tree, enables the unmanned vessel to adaptively switch between source tracing, obstacle avoidance, and return to shore, thereby ensuring navigation safety and mission continuity. In this way, the overall intelligence level and engineering applicability of the unmanned vessel system are improved, which is conducive to the deployment, operation, and promotion of this water pollution source tracing method in actual aquatic environments.

[0079] The above are preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions made by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for tracing the source of water pollution from unmanned vessels, characterized in that, Includes the following steps: S1. Initiate the pollution source tracing task; S2. Construct an improved dynamic behavior tree decision model, including three parallel behavior subtrees: source tracing, obstacle avoidance, and return to home. S3. Real-time collection of pollutant concentration, obstacle distance, and remaining battery power in four directions (front, rear, left, and right) of the unmanned vessel; S4. Based on the environmental data collected in S3, the unmanned vessel calculates the source tracing weight, obstacle avoidance weight, and return weight in each sampling period T based on the improved dynamic behavior tree decision model, and performs normalization processing. S5. Based on the behavior switching logic of the improved dynamic behavior tree decision model, a judgment is made. If the behavior switching logic is satisfied, the behavior subtree with the highest weight is executed; otherwise, the current behavior continues to be executed. S6. Based on the final decision in S5, the unmanned vessel executes the control algorithm corresponding to the behavior subtree to complete the corresponding action.

2. The method for tracing the source of water pollution from unmanned vessels as described in claim 1, characterized in that, The improved dynamic behavior tree decision model includes: a root node (101), behavior subtree nodes (102), and action nodes (103). The root node (101) calculates the normalized weight values ​​of the source tracing, obstacle avoidance, and return navigation behavior subtrees based on the environmental data. When the normalized weight value of a non-current behavior satisfies the behavior switching logic, a behavior switch is triggered, thereby activating a behavior subtree in the next layer. The behavior subtree nodes (102) include parallel source tracing behavior subtrees, obstacle avoidance behavior subtrees, and return navigation behavior subtrees. The entry point of each subtree is a sequential node, used to sequentially trigger the action nodes (103) under that subtree. The action nodes (103) are terminal nodes, used to execute the control algorithm corresponding to the action.

3. A method for tracing the source of water pollution from unmanned vessels as described in claim 1 or 2, characterized in that, The behavior switching logic described in S5 includes the following two constraints: S51. Hysteresis constraints require that the normalized weight value of the behavior to be switched must be higher than the sum of the normalized weight value of the currently executing behavior and the set threshold increment Δ; S52. The time constraint requires that, under the premise of satisfying the hysteresis constraint, the duration of this weighting relationship exceeds the set anti-shake time T. switch .

4. A system for tracing the source of water pollution from unmanned vessels, characterized in that, include: The unmanned vessel body, and the main control module (1), concentration detection module (2), obstacle avoidance detection module (3), power monitoring module (4), positioning module (5), drive module (6) and wireless communication module (7) installed on the unmanned vessel body; also includes a host computer control module (8) that is connected to the wireless communication module; Among them, the concentration detection module (2) is used to obtain pollutant concentration data in the four directions of front, back, left and right of the unmanned vessel; the obstacle avoidance detection module (3) is used to obtain obstacle distance data in the four directions of front, back, left and right of the unmanned vessel; the power monitoring module (4) is used to obtain the remaining power data of the unmanned vessel; the positioning module (5) is used to obtain the location information of the unmanned vessel; all of the above modules are connected to the main control module (1) for data. The main control module (1) stores and runs the improved dynamic behavior tree decision model, which includes three parallel behavior subtrees: source tracing, obstacle avoidance, and return. The main control module (1) calculates the source tracing weight, obstacle avoidance weight and return weight based on pollutant concentration data, obstacle distance data and remaining power data in each sampling period T and performs normalization processing. It then determines the behavior subtree to be executed according to the preset behavior switching logic and outputs control commands to the drive module (6) to drive the unmanned vessel to perform the corresponding actions. The host computer control module (8) is used for task initialization, remote monitoring, environmental data visualization and trajectory display. The navigation behavior decision and control of the unmanned ship is completed by the main control module (1).

Images