Path planning method and system for water quality detection unmanned ship in autonomous cruise

Abstract

The application relates to a water quality detection unmanned ship path planning method and system of autonomous cruise, relates to the ship path planning technical field, and comprises the following steps: determining the monitoring frequency of each monitoring point based on the historical prior knowledge of each monitoring point in a target water area, and generating a monitoring list cycle according to the monitoring frequency; in response to a monitoring task starting instruction, extracting a real-time monitoring list to be executed at present from the monitoring list cycle; obtaining a dynamic cost field representing the water environment resistance in the target water area, and based on a multi-objective optimization method, combining the real-time monitoring list and the dynamic cost field to perform global path optimization, and obtaining a globally optimized path; and delivering the globally optimized path to the unmanned ship to perform local navigation and adaptive adjustment. The application solves the problems that the traditional ship path planning method is difficult to accurately consider the complex environmental changes of the water area, the planned path cannot meet the actual navigation requirements, energy consumption is high, efficiency is low, and an effective adaptive adjustment mechanism is lacked when facing unexpected conditions.

Description

Technical Field

[0001] This application relates to the field of vessel path planning, and in particular to a method and system for path planning of unmanned surface vessels (USVs) for autonomous water quality monitoring. Background Technology

[0002] With the increasing demands for detection efficiency, data timeliness, and endurance of unmanned equipment in the field of water quality monitoring, the rationality of path planning for autonomous cruise water quality monitoring unmanned vessels has become a key technical requirement to ensure the efficient implementation of monitoring tasks.

[0003] Currently, traditional ship path planning methods do not fully consider the complex changes in the aquatic environment and cannot specifically adapt to the impact of dynamic factors such as water flow and resistance on navigation. This not only easily leads to a disconnect between the planned path and the actual navigation needs, resulting in high energy consumption and low monitoring efficiency of unmanned vessels, but also increases the risk of monitoring mission interruption or incomplete data collection due to the lack of an effective adaptive adjustment mechanism in the face of emergencies. Summary of the Invention

[0004] This application provides a method and system for autonomous navigation of unmanned surface vessels (USVs) for water quality monitoring, which improves the adaptability of USV paths to complex changes in aquatic environments, addresses the disconnect between planned paths and actual navigation needs, reduces navigation energy consumption, and enhances monitoring efficiency.

[0005] The embodiments of this application disclose the following technical solutions:

[0006] In a first aspect, embodiments of this application provide a path planning method for autonomous cruise unmanned surface vessel (USV) for water quality testing, the method comprising:

[0007] Based on the historical prior knowledge of each monitoring point in the target water area, the monitoring frequency of each monitoring point is determined, and a monitoring list is generated cyclically according to the monitoring frequency.

[0008] In response to the monitoring task start command, the current real-time monitoring list to be executed is extracted from the monitoring list in a loop;

[0009] A dynamic cost field characterizing the water environment resistance within the target water area is obtained, and a global path optimization is performed based on a multi-objective optimization method, combining the real-time monitoring list with the dynamic cost field, to obtain a globally optimized path.

[0010] The globally optimized path is sent to the unmanned vessel for local navigation and adaptive adjustment.

[0011] Secondly, embodiments of this application provide an autonomous navigation unmanned surface vessel (USV) path planning system for water quality monitoring, the system comprising:

[0012] The monitoring frequency and inventory generation module is used to determine the monitoring frequency of each monitoring point based on the historical prior knowledge of each monitoring point in the target water area, and generate a monitoring inventory cyclically according to the monitoring frequency.

[0013] The real-time monitoring list extraction module is used to respond to the monitoring task start command and extract the real-time monitoring list that needs to be executed from the monitoring list in a loop.

[0014] The cost field and path optimization module is used to obtain a dynamic cost field that characterizes the water environment resistance in the target water area, and to perform global path optimization based on a multi-objective optimization method, combined with the real-time monitoring list and the dynamic cost field, to obtain a globally optimized path.

[0015] The path distribution and navigation module is used to distribute the globally optimized path to the unmanned vessel for local navigation and adaptive adjustment.

[0016] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0017] This application proposes an autonomous navigation unmanned surface vessel (USV) path planning method and system for water quality monitoring. By determining the monitoring frequency and monitoring list cycle step by step, extracting the real-time monitoring list, constructing a dynamic cost field and global path optimization, and performing local navigation and adaptive adjustment, the USV achieves efficient and low-consumption water quality monitoring path planning in target waters. First, based on historical prior knowledge of each monitoring point in the target water area, the water quality fluctuation coefficient is calculated and weighted with the location importance weight to obtain a priority score. Combined with the monitoring frequency mapping table, the monitoring frequency of each point is determined. Then, the least common multiple of the monitoring frequencies is calculated to set the list cycle, the task time slices are divided and the sampling timestamps are distributed to form a monitoring list cycle. Next, in response to the monitoring task start command, the current real-time monitoring list is extracted from the list cycle. Subsequently, historical flow velocity and flow direction data of the target water area are obtained to generate a flow velocity-flow direction distribution map. The distribution of relative resistance coefficients is calculated to construct a dynamic cost field. Using the dynamic cost field as the energy consumption basis and the point priority score as the information benefit metric, a multi-objective optimization function is constructed. The global optimization path is solved by combining the unmanned surface vessel's endurance constraint and path optimization algorithm. Finally, the global optimization path is issued to the unmanned surface vessel, controlling it to navigate along the path and collect water quality data. During navigation, the path is adaptively adjusted in real time based on the energy consumption rate and endurance model. At the same time, the flow velocity and flow direction are inverted and the dynamic cost field is updated based on the navigation log and adjustment records, providing accurate environmental parameters for subsequent path planning.

[0018] The technical solution proposed in this application solves the problems in traditional ship path planning, such as difficulty in accurately considering complex changes in the aquatic environment, high energy consumption and low efficiency due to the disconnect between the planned path and actual navigation needs, and lack of effective adaptive adjustment mechanisms in the face of emergencies. It avoids insufficient endurance, interruption of monitoring tasks or incomplete data collection caused by deviations in environmental resistance estimation or path rigidity, and improves the adaptability, economy and reliability of path planning for autonomous cruise water quality detection unmanned ships. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating the path planning method for autonomous navigation water quality detection unmanned surface vessel provided in this application embodiment;

[0021] Figure 2 A schematic diagram of the path planning system for autonomous cruise water quality detection unmanned surface vessel provided in this application embodiment.

[0022] The components represented by each number in the attached diagram are explained below:

[0023] Monitoring frequency and list generation module 01, real-time monitoring list extraction module 02, cost field and path optimization module 03, path distribution and navigation module 04. Detailed Implementation

[0024] This application provides a method and system for autonomous navigation water quality detection unmanned vessel path planning, which addresses the technical problems in existing technologies, such as the difficulty in accurately considering complex changes in the aquatic environment during vessel path planning, resulting in planned paths that cannot meet actual navigation needs, high energy consumption, low efficiency, and a lack of effective adaptive adjustment mechanisms in the face of emergencies.

[0025] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0027] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0028] Example 1, as shown in the appendix Figure 1 As shown, this application provides a path planning method for autonomous cruise unmanned surface vessel (USV) for water quality monitoring, the method comprising the following steps:

[0029] S110: Based on the historical prior knowledge of each monitoring point in the target water area, determine the monitoring frequency of each monitoring point, and generate a monitoring list cyclically according to the monitoring frequency;

[0030] In this embodiment of the application, in the scenario where an autonomous cruise water quality detection unmanned surface vessel (USV) performs water monitoring tasks, in order for the USV to accurately cover each monitoring point according to differentiated needs, it is necessary to first determine the monitoring frequency of each point based on historical prior knowledge, and then construct an orderly monitoring list cycle accordingly, so as to improve the monitoring efficiency and endurance resource utilization of the USV.

[0031] Specifically, the first step is to acquire historical prior knowledge of each monitoring point within the target water area. This knowledge includes historical water quality monitoring data and location importance weights for the target area. The former reflects past changes in water quality at the point, while the latter reflects the importance of the point in water area monitoring.

[0032] Furthermore, the water quality fluctuation coefficient is calculated based on the acquired historical water quality monitoring data. The mean square error method is used to quantify the degree of fluctuation of water quality data during the calculation process. Then, the water quality fluctuation coefficient is weighted and fused with the location importance weight. The priority score of each monitoring point is obtained through the fusion calculation. The higher the score, the more frequently monitoring needs to be carried out at that point.

[0033] Furthermore, by combining the preset monitoring frequency mapping table, the numerical range to which the priority score of each monitoring point belongs is determined, and the monitoring frequency level corresponding to each monitoring point is determined based on the interval correspondence, thereby clarifying the specific monitoring frequency.

[0034] Furthermore, the least common multiple of the monitoring frequencies of multiple monitoring points is calculated, and the duration corresponding to this least common multiple is used as the list cycle to ensure that all points can complete monitoring at their own frequencies within the cycle.

[0035] Meanwhile, taking into account the typical endurance characteristics of unmanned vessels, the inventory cycle is divided into multiple continuous and fixed-duration task time slices to ensure that the navigation and monitoring tasks of the unmanned vessels can be completed within the endurance range in each time slice.

[0036] Furthermore, based on the monitoring frequency of multiple monitoring points, sampling timestamps are distributed within the list cycle to clarify the specific time nodes at which each point needs to be monitored. For each task time slice, based on the sampling timestamp distribution results, multiple monitoring points are traversed to determine the sampling object and whether a certain point needs to be monitored within the time slice. If monitoring is required, the point is added to the monitoring list of the corresponding task time slice, ultimately forming a monitoring list cycle. This monitoring list cycle contains multiple monitoring lists arranged in chronological order, and each monitoring list defines the set of monitoring points that need to be accessed in a monitoring task.

[0037] This step uses historical prior knowledge to determine the monitoring frequency and generate a monitoring list cycle, providing a clear path guide for the subsequent unmanned surface vessel to perform monitoring tasks in sequence. This enables the unmanned surface vessel to prioritize the coverage of high-priority locations within its limited range, while ensuring the orderly implementation of monitoring needs for all locations.

[0038] Step S110 in the method provided in this application embodiment includes:

[0039] The historical prior knowledge is obtained, wherein the historical prior knowledge includes historical water quality monitoring data and location importance weights for the target area;

[0040] The water quality fluctuation coefficient is calculated based on the historical water quality monitoring data, and the water quality fluctuation coefficient is weighted and fused with the location importance weight to obtain the priority score of each monitoring point.

[0041] By combining the preset monitoring frequency mapping table, the numerical range to which the priority score belongs is determined, and the monitoring frequency level and the corresponding monitoring frequency for each monitoring point are determined.

[0042] Calculate the least common multiple of the monitoring frequencies of multiple monitoring points, and use the duration corresponding to the least common multiple of the frequencies as the list cycle period;

[0043] Based on the typical endurance characteristics of unmanned vessels, the inventory cycle is divided into multiple consecutive task time slices with fixed durations.

[0044] Based on the monitoring frequency of multiple monitoring points, the sampling timestamps are distributed in the inventory cycle.

[0045] For each task time slice, based on the sampling timestamp distribution results, multiple monitoring points are traversed to identify the sampling objects, and the monitoring points are added to the monitoring list corresponding to each task time slice to form the monitoring list loop.

[0046] In this embodiment of the application, in order to enable the autonomous water quality monitoring unmanned surface vessel to carry out monitoring tasks in a targeted manner and avoid the waste of energy and the loss of key data points caused by indiscriminate monitoring of all monitoring points, it is necessary to first determine the monitoring frequency of each point based on historical prior knowledge, and then build an orderly monitoring list cycle around the frequency to ensure that the unmanned surface vessel can efficiently complete water quality monitoring within its limited range.

[0047] Specifically, the first step is to obtain the historical prior knowledge of each monitoring point within the target area, which includes two types of core information: one is the historical water quality monitoring data of the target area, and the other is the locational importance weight of each monitoring point.

[0048] Historical water quality monitoring data must cover continuous records within a preset time period, such as pH values ​​and pollutant content data from the past 1-3 years, to ensure a complete reflection of the past changes in water quality at each location. The weight of location importance must be determined in conjunction with the actual functional attributes of the location. For example, the weight of locations around drinking water sources is set at 0.7-0.9, and the weight of ordinary landscape river locations is set at 0.3-0.5, to reflect the differences in the importance of different locations in water quality monitoring.

[0049] Furthermore, the water quality fluctuation coefficient is calculated using the mean square error method based on historical water quality monitoring data. For example, the mean square error is used to calculate the dispersion of dissolved oxygen concentration at a certain point over the past 12 months. The greater the dispersion, the higher the water quality fluctuation coefficient, indicating that the water quality at that point changes more frequently and requires closer monitoring.

[0050] Furthermore, the water quality fluctuation coefficient and the location importance weight are weighted and integrated. During the integration, the weight ratio of the two can be set according to actual needs. For example, the water quality fluctuation coefficient accounts for 0.6 and the location importance weight accounts for 0.4. The priority score of each monitoring point is obtained through weighted calculation. The higher the score, the higher the monitoring priority that the point should be given.

[0051] Furthermore, by combining the preset monitoring frequency mapping table, the numerical range to which the priority score of each monitoring point belongs is determined.

[0052] The monitoring frequency mapping table needs to be set in advance according to the monitoring task requirements. For example, a priority score of 80-100 corresponds to frequency level 1, with a monitoring frequency of once a day; a score of 60-79 corresponds to frequency level 2, with a monitoring frequency of once every 2 days; a score of 40-59 corresponds to frequency level 3, with a monitoring frequency of once every 3 days; and a score below 40 corresponds to frequency level 4, with a monitoring frequency of once a week. Based on the above interval correspondence, the monitoring frequency level and specific monitoring frequency corresponding to each monitoring point are determined.

[0053] Furthermore, calculate the least common multiple of the monitoring frequencies of multiple monitoring points. For example, if there are 3 monitoring points in a certain water area, with monitoring frequencies of once a day, once every 2 days, and once every 3 days, and the least common multiple of the three is 6, then use 6 days as the list cycle to ensure that within the 6-day cycle, the 3 points can complete monitoring at frequencies of once / day, once / 2 days, and once / 3 days respectively, without missing the monitoring needs of any point.

[0054] Furthermore, considering the typical endurance characteristics of unmanned vessels, the inventory cycle is divided into multiple continuous and fixed-duration task time slices.

[0055] Among them, the typical endurance characteristics of unmanned ships need to be determined based on actual navigation tests. For example, if a certain model of unmanned ship can support up to 4 hours of navigation and monitoring on a single full charge, then the 6-day inventory cycle (a total of 144 hours) is divided into 36 task time slices of 4 hours each, to ensure that the unmanned ship can complete the corresponding monitoring task within the endurance range in each time slice, and to avoid mission interruption due to insufficient endurance.

[0056] Furthermore, based on the monitoring frequency of multiple monitoring points, sampling timestamps are distributed within the inventory cycle. For example, for a point monitored once a day, six sampling timestamps need to be distributed within a 6-day cycle, and the timestamp intervals need to be kept uniform, such as 9:00 AM every day; for a point monitored once every two days, three sampling timestamps are distributed within a 6-day cycle, such as 9:00 AM on the 2nd, 4th, and 6th days; for a point monitored once every three days, two sampling timestamps are distributed within a 6-day cycle, such as 9:00 AM on the 3rd and 6th days. Through the above distribution method, it is ensured that the sampling time of each point meets its own monitoring frequency requirements.

[0057] Finally, for each task time slice, based on the sampling timestamp distribution results, multiple monitoring points are traversed to identify the sampling objects. For example, for the task time slice "9:00 AM - 1:00 PM on the second day", it is checked whether there is a sampling timestamp for a certain point within this time slice. If a point with a monitoring frequency of once every two days has a sampling timestamp within this time slice, then this point is determined to be a sampling object for this time slice and added to the monitoring list corresponding to this task time slice.

[0058] Following the same method, the sampling object identification and point addition for all task time slices are completed, and finally a monitoring list loop is formed. This monitoring list loop contains 36 monitoring lists arranged in chronological order. Each monitoring list clearly defines the set of monitoring points that the unmanned vessel needs to visit within the corresponding 4-hour task time slice.

[0059] S120: In response to the monitoring task start command, extract the current real-time monitoring list to be executed from the monitoring list in a loop;

[0060] In this embodiment of the application, in order to make the water quality monitoring unmanned vessel know the specific target location for each monitoring task, it is necessary to extract the monitoring point list for the corresponding time period from the generated monitoring list in a loop when the monitoring task is started, so as to ensure that the unmanned vessel can quickly meet the task requirements and efficiently carry out water quality data collection.

[0061] Specifically, the first step is to establish a response mechanism for the monitoring task initiation command. This mechanism needs to be linked with the unmanned vessel's task scheduling platform. When the task scheduling platform issues a monitoring task initiation command, the unmanned vessel will automatically trigger the inventory retrieval process.

[0062] The monitoring task initiation command must include the task execution time information. For example, the command may specify "execute the monitoring task for the 3rd task time slice". This time information will serve as the core basis for extracting the real-time monitoring list from the monitoring list in a loop, ensuring that the extracted list is completely matched with the current task time period.

[0063] Furthermore, the task scheduling platform will retrieve the generated monitoring list loop and locate the task time slice corresponding to the task to be executed based on the time information in the monitoring task start instruction. For example, if the instruction points to "the 5th 4-hour task time slice within the list loop cycle", the platform will find the monitoring list entry corresponding to that time slice in the list loop.

[0064] During the extraction process, the information of the located monitoring list items needs to be verified. The verification includes whether the monitoring point names, coordinates, and planned monitoring sequence are complete. At the same time, the extracted monitoring list needs to be initially adapted to the navigation platform of the unmanned surface vessel (USV) to confirm that the coordinate format of all points in the list can be recognized by the navigation platform, ensuring that the USV can successfully plan its navigation path according to the list.

[0065] If the verification is successful, the platform will identify the monitoring list as the current real-time monitoring list to be executed and send it to the control terminal of the unmanned vessel. If the verification finds that the list information is missing or the format is mismatched, the platform will automatically return to the list loop database to extract it again until a complete and suitable real-time monitoring list is extracted.

[0066] S130: Obtain the dynamic cost field characterizing the water environment resistance in the target water area, and based on the multi-objective optimization method, combine the real-time monitoring list with the dynamic cost field to perform global path optimization and obtain the global optimized path;

[0067] In this embodiment of the application, in order to make the navigation path of the water quality monitoring unmanned vessel adaptable to changes in water environment resistance and balance energy consumption and monitoring efficiency, it is necessary to first construct a dynamic cost field reflecting water environment resistance, and then combine a multi-objective optimization method to plan the optimal path, so as to ensure that the unmanned vessel completes the monitoring task efficiently and with low energy consumption.

[0068] Specifically, the historical flow velocity and flow direction data of the target water area are first obtained, and a flow velocity-flow direction distribution map of the target water area is generated based on the above two types of data to clearly present the water flow status of each area within the water area.

[0069] Furthermore, using the baseline navigation resistance in static water as a reference, and combining it with the velocity-direction distribution diagram, the relative resistance coefficient distribution of multiple points at equal preset intervals in the target water area is calculated. During the calculation process, the resistance differences of each point in multiple navigation directions with equal preset azimuth angle differences need to be considered, and the relative resistance coefficient in each direction is calculated separately, ultimately forming a dynamic cost field with anisotropic characteristics.

[0070] Furthermore, using the dynamic cost field as the basis for path energy consumption calculation, and the priority score corresponding to each monitoring point in the real-time monitoring list as the metric for information collection benefits, an optimization function is constructed with minimizing total path energy consumption, maximizing path energy efficiency, and maximizing path information collection efficiency as the core optimization objectives.

[0071] Meanwhile, the optimization function is solved by combining the unmanned vessel's endurance constraints and path optimization algorithms. The algorithm calculates the optimal path sequence for accessing all points in the real-time monitoring list. This sequence must simultaneously meet the requirements of energy consumption, efficiency, and endurance. Finally, the solved path sequence is output as the global optimized path.

[0072] This step, by constructing a dynamic cost field and multi-objective optimized path, provides unmanned surface vessels with navigation guidance that fits the actual water environment and takes into account multiple needs. It avoids the excessive impact of water flow resistance on navigation energy consumption, while ensuring the efficiency and data value of monitoring tasks, laying the foundation for the subsequent accurate execution of local navigation by unmanned surface vessels.

[0073] Step S130 in the method provided in this application embodiment includes:

[0074] Obtain historical flow velocity and direction data for the target water area, and generate a flow velocity-direction distribution map of the target water area;

[0075] Using the baseline navigation resistance in a static water body as a reference, and based on the velocity-direction distribution diagram, the relative resistance coefficient distribution of multiple points at a predetermined interval in the target water area is calculated to form the dynamic cost field.

[0076] The relative drag coefficient distribution includes multiple relative drag coefficients in multiple navigation directions with equal preset azimuth angle differences, and the relative drag coefficient distribution at each point is anisotropic.

[0077] The dynamic cost field is used as the basis for path energy consumption calculation;

[0078] The priority score corresponding to each monitoring point in the real-time monitoring list is used as a measure of information collection benefit.

[0079] An optimization function is constructed with the objectives of minimizing total path energy consumption and maximizing path energy efficiency and path information collection efficiency. The optimal path sequence for accessing the real-time monitoring list is solved by combining the unmanned vessel endurance constraint and the path optimization algorithm, and the output is the global optimized path.

[0080] In this embodiment of the application, in order to enable the water quality monitoring unmanned surface vessel to accurately adapt to the changes in water environment resistance of the target water area and efficiently complete the monitoring of the points in the real-time monitoring list within a limited range, it is necessary to first construct a dynamic cost field that reflects the differences in water environment resistance, and then plan the global path with multi-objective optimization as the core, so as to ensure the economy and effectiveness of the unmanned surface vessel's navigation and monitoring.

[0081] Specifically, the first step is to acquire historical flow velocity and direction data for the target water area. This data needs to cover the flow characteristics of different time periods and regions of the target water area, such as flow velocity records during the high-water season, low-water season, and normal-water season, as well as flow direction differences in different areas such as nearshore shallow water areas, deep water areas in the center of the lake, and river confluences, to ensure that the data can comprehensively reflect the dynamic changes in the flow of the target water area.

[0082] Furthermore, through data visualization and spatial interpolation, a velocity-direction distribution map of the target water area is generated. This distribution map needs to clearly indicate the velocity range and direction of flow in each area, providing an intuitive spatial reference for subsequent resistance calculations.

[0083] Furthermore, using the baseline navigation resistance in static water as a reference, the relative resistance coefficient distribution of multiple points at predetermined intervals in the target water area is calculated based on the velocity-direction distribution diagram.

[0084] The setting of the preset spacing needs to be determined in combination with the target water area and the monitoring accuracy requirements. For example, a 50-meter spacing is set when the water area is small and a 100-meter spacing is set when the area is large, to ensure the uniformity and representativeness of the point coverage.

[0085] When calculating the relative drag coefficient, it is necessary to divide each point into multiple navigation directions according to a preset azimuth angle difference, for example, dividing it into 12 directions at 30° intervals, and calculating the drag effect of water flow on the unmanned vessel navigation in each direction to obtain the relative drag coefficient of each point in different directions. Finally, an anisotropic dynamic cost field is formed, that is, the drag coefficient of the same point is different in different navigation directions, so as to conform to the influence law of water flow direction on navigation resistance in actual waters.

[0086] The method provided in this application embodiment, after "obtaining the dynamic cost field characterizing the water environment resistance in the target water area", further includes:

[0087] Retrieve the historical path scheme corresponding to the real-time monitoring list from the previous monitoring list loop;

[0088] Verifying the feasibility of the historical path scheme by combining the dynamic cost field includes:

[0089] Verify the geographical feasibility of the historical route scheme and determine whether the maximum point of the relative resistance coefficient of the historical route scheme under the dynamic cost field meets the preset maximum resistance coefficient threshold.

[0090] If the geographical feasibility verification is successful, the endurance feasibility of the historical route plan is further verified. The total energy consumption of traveling along the historical route plan is estimated based on the dynamic cost field, and it is determined whether the total energy consumption is within the preset safe endurance threshold.

[0091] If both geographical feasibility and range feasibility are verified, the historical route scheme will be adopted as the global optimized route.

[0092] If either geographical feasibility or range feasibility fails the verification, then global path optimization will be performed.

[0093] In this embodiment of the application, in order to ensure that the path of the water quality detection unmanned vessel is adapted to the current water environment and reduce unnecessary path calculation overhead, it is necessary to first retrieve the historical path schemes of the corresponding list from the previous cycle and verify their feasibility in combination with the current dynamic cost field, so that the path can be reused directly when it is available and the optimization can be restarted when it is unavailable, thereby improving the efficiency and economy of path planning.

[0094] Specifically, the first step is to retrieve the historical path scheme from the previous monitoring list loop that corresponds to the current real-time monitoring list. This "correspondence" must be based on matching the core features of the monitoring list. For example, if the current real-time monitoring list contains three high-priority points A, B, and C, and the task time slice is 4 hours, then the historical path scheme from the previous loop that also contains these three points and has the same task time slice duration must be retrieved. This ensures that the monitoring objectives and execution conditions of the two are comparable, preventing the reused path from failing to meet the current task requirements due to differences in list content.

[0095] Furthermore, the feasibility of the historical route scheme was verified by combining the obtained dynamic cost field. The verification process was carried out in two steps: geographical feasibility and endurance feasibility.

[0096] First, geographical feasibility is verified to determine whether there are areas where unmanned vessels cannot pass under the current water environment resistance of historical route schemes. Specifically, all navigation segments covered by historical route schemes need to be extracted, and the maximum relative resistance coefficient of each segment needs to be found in the dynamic cost field (i.e., the location with the greatest resistance in the path).

[0097] If the maximum value of the relative resistance coefficient does not exceed the preset maximum resistance coefficient threshold (set according to the power output capability of the unmanned vessel's propulsion device, the vessel's navigation performance, and the common resistance characteristics of the target water area), it means that the historical route plan has no geographical obstacles in the current water environment, and the geographical feasibility verification is passed; if the maximum value of the relative resistance coefficient exceeds the maximum resistance coefficient threshold, it means that there are high resistance areas in the historical route plan that the unmanned vessel cannot pass through, and the geographical feasibility verification is not passed.

[0098] Furthermore, if the geographical feasibility verification is successful, the endurance feasibility verification will proceed to determine whether the total energy consumption of sailing along the historical route is within the safe endurance range of the unmanned vessel.

[0099] Specifically, based on the distribution of relative drag coefficients in the dynamic cost field, and combined with the energy consumption model of the unmanned vessel, the energy consumption of each segment in the historical path is calculated segment by segment, and then the energy consumption of all segments is summed to obtain the total energy consumption for sailing along the path.

[0100] Subsequently, the total energy consumption is compared with the preset safe endurance threshold (the safe endurance threshold is usually set at 80% of the total energy consumption of the unmanned vessel when fully charged, with 20% reserved for emergencies). If the total energy consumption is less than or equal to the safe endurance threshold, the endurance feasibility verification is passed; if the total energy consumption exceeds the safe endurance threshold, it indicates that sailing along the historical route may lead to insufficient endurance of the unmanned vessel, and the endurance feasibility verification fails.

[0101] Furthermore, if both geographical feasibility and endurance feasibility are verified, it indicates that the historical route plan is adaptable to the current water environment and the state of the unmanned vessel. This plan can be directly adopted as the global optimized route for the current mission without recalculation, effectively shortening the route planning time.

[0102] Conversely, if any one of the feasibility measures fails verification, such as failing geographical feasibility due to high resistance areas, or failing range feasibility due to excessive energy consumption, then the reuse of historical path solutions must be abandoned, and global path optimization must be initiated to find a new optimal path.

[0103] Furthermore, when performing global path optimization, the dynamic cost field is first used as the basis for path energy consumption calculation, and the priority score corresponding to each monitoring point in the real-time monitoring list is used as the information collection benefit metric. An optimization function is constructed with the goal of minimizing total path energy consumption and maximizing path energy consumption efficiency and path information collection efficiency.

[0104] Among them, path energy efficiency is measured by the ratio of the actual planned path length to the maximum range path length of the unmanned vessel. The calculation formula can be expressed as "path energy efficiency = actual planned path length / maximum range path length of the unmanned vessel". The closer the ratio is to 1, the more fully the path utilizes the range resources.

[0105] In addition, the efficiency of path information collection is measured by the ratio of the sum of priority scores of all monitoring points covered by the path to the geometric length of the path. The calculation formula can be expressed as "path information collection efficiency = sum of priority scores of all monitoring points / geometric length of the path". The higher the ratio, the greater the monitoring value obtained per unit path length.

[0106] At the same time, when constructing the optimization function, it is also necessary to take into account the unmanned vessel's range constraint, that is, the total energy consumption of the planned path must not exceed the total energy consumption of the unmanned vessel when it is fully charged, so as to avoid the unmanned vessel being unable to complete the mission or returning due to energy consumption exceeding the limit.

[0107] Furthermore, a path optimization algorithm is used to solve for the optimal path sequence to access the real-time monitoring list. Specifically, a greedy algorithm from existing technologies is adopted. Starting from the current starting position, at each step, the monitoring point with the minimum energy consumption increment and maximum information gain increment is selected as the next station, and all points are traversed step by step.

[0108] During the selection process, the energy consumption increment to each candidate point needs to be calculated based on the dynamic cost field, which is the distance from the current position to the candidate point. Combined with the relative resistance coefficient corresponding to the route, the energy consumption increment is calculated through the existing unmanned vessel energy consumption model.

[0109] Meanwhile, the incremental information gain is determined based on the priority score of the candidate points; the higher the priority score, the greater the incremental information gain. If multiple candidate points have similar overall performance in terms of energy consumption increment and information gain increment, the matching degree between the navigation direction and the water flow direction in the dynamic cost field is further considered, and points in the downstream or lateral downstream direction are prioritized to further reduce energy consumption during actual navigation.

[0110] Finally, based on the above step-by-step selection method, the access order of all monitoring points is determined in sequence, and a complete path sequence is formed, which is the optimal path sequence. This ensures that when the unmanned vessel travels along this path, it can minimize the total energy consumption and improve the energy utilization efficiency, while also prioritizing the coverage of high-priority monitoring points to ensure the information collection benefits, and at the same time meeting the unmanned vessel's endurance constraints.

[0111] S140: Send the global optimized path to the unmanned vessel for local navigation and adaptive adjustment.

[0112] In this embodiment of the application, in order to ensure that the autonomous cruise water quality monitoring unmanned vessel can complete the water quality monitoring task as planned, and to cope with the energy consumption fluctuations and path deviations caused by changes in the water environment during navigation, the global optimized path needs to be sent to the unmanned vessel to guide it to carry out local navigation, real-time adaptive adjustment and dynamic cost field update, so as to ensure the effectiveness and accuracy of the monitoring task.

[0113] Specifically, the unmanned surface vessel (USV) is first controlled to navigate along a globally optimized path and collect water quality data. The USV's control terminal uses the received globally optimized path as a navigation reference and combines it with real-time positioning data to correct its course and speed, ensuring that the vessel travels along the preset path. When it reaches each monitoring point, it automatically activates the water quality detection equipment, collects key water quality parameters such as pH value, and stores the data in association with the location information.

[0114] Furthermore, during navigation, the global optimized path is adaptively adjusted in real time based on the energy consumption rate and the local endurance estimation model. The unmanned surface vessel continuously monitors energy consumption and adjusts its speed or avoids high-energy-consuming areas in a timely manner, ensuring that the mission can be completed within its endurance range without missing any monitoring points. If it encounters sudden changes in water flow, obstacles, or other emergencies, it will also ensure navigation safety through local path adjustments.

[0115] Simultaneously, the dynamic cost field is updated based on local navigation logs and adaptive adjustment records. By organizing and analyzing positioning, attitude, energy consumption, and adjustment data during navigation, the actual flow velocity and direction in the water are deduced. Then, the relative drag coefficient distribution of the dynamic cost field is updated in conjunction with real-time data, so that the generated optimized path can more accurately adapt to the current water environment and improve the overall efficiency of unmanned surface vessel path planning.

[0116] This step combines path distribution, navigation execution, real-time adjustment, and dynamic cost field updates to achieve efficient implementation of the current monitoring task while collecting accurate environmental data for path planning, effectively improving the reliability and safety of unmanned vessel water quality monitoring.

[0117] Step S140 in the method provided in this application embodiment includes:

[0118] Control the unmanned vessel to navigate along the globally optimized path and collect water quality data;

[0119] During the voyage, the global optimization path is adaptively adjusted in real time based on the energy consumption rate and the local range estimation model;

[0120] The dynamic cost field is updated accordingly based on the local navigation logs and adaptive adjustment records.

[0121] In this embodiment of the application, in order to ensure that the unmanned vessel can accurately perform monitoring tasks according to the path, and to cope with possible changes in the water environment or emergencies during navigation, the globally optimized path needs to be sent to the unmanned vessel control terminal, and the effective execution of the path is achieved by relying on the local navigation device and dynamic adjustment mechanism, so as to ensure the continuity of monitoring tasks, the integrity of data collection and the navigation safety of the unmanned vessel.

[0122] Specifically, the first step is to establish a global optimization path distribution and reception mechanism. This mechanism requires a stable data transmission link between the path planning platform and the unmanned vessel control terminal, such as through 4G / 5G wireless communication or satellite communication, to avoid the loss of path information due to transmission interruption.

[0123] Before the data is distributed, the global optimized path data needs to be formatted. The coordinate sequence, point access order, and estimated travel time output by the path planning platform should be converted into an instruction format that the unmanned vessel control terminal can recognize, so as to ensure that the unmanned vessel can accurately parse the path content.

[0124] Furthermore, the unmanned surface vessel activates its local navigation system to execute a path-based navigation. The local navigation system uses the received globally optimized path as a basis, combined with real-time acquired GPS positioning data and IMU inertial measurement data, to correct the navigation attitude and position in real time.

[0125] For example, when the unmanned vessel deviates from the planned path due to water flow disturbance, the navigation device will calculate and adjust the angle and propulsion based on the positioning deviation to control the vessel to return to the preset path; at the same time, when it sails to each monitoring point, the navigation device will trigger the water quality detection equipment to start, collect data such as pH value and pollutant concentration at a preset sampling frequency, and transmit the collection results back to the background monitoring platform in real time.

[0126] During local navigation, the unmanned vessel will also make adaptive adjustments to cope with changes in the aquatic environment or emergencies.

[0127] Specifically, the water quality detection unmanned vessel first records the real-time energy consumption rate through an energy consumption monitoring device. Combined with the remaining power and range estimation model, it determines whether there is an energy consumption exceeding expectations on the current path. If the energy consumption of a certain section increases too quickly, it may mean that the actual water flow resistance in that area is greater than the preset value of the dynamic cost field. At this time, the navigation speed will be adjusted appropriately or the high energy consumption area will be bypassed to ensure that the total energy consumption is controlled within a safe range.

[0128] Secondly, if the water quality of the current water area is detected in real time by the water quality sensor, or if the obstacle in the path is detected by the obstacle avoidance sensor, a temporary path adjustment will be triggered. Without deviating from the overall monitoring target, the local road segment will be replanned, and the path will be returned to the global optimization after the abnormality or obstacle is eliminated.

[0129] Meanwhile, during the adaptive adjustment process, the unmanned vessel will record adjustment logs in real time, including the reason for the adjustment, the path changes before and after the adjustment, the energy consumption and navigation status after the adjustment, and other information. The logs will be synchronously transmitted back to the path planning platform to provide a reference for subsequent path optimization of this mission.

[0130] In addition, if a major anomaly occurs during local navigation and adaptive adjustment, such as the unmanned vessel encountering strong currents and being unable to autonomously return to its path, or having insufficient remaining power to complete the remaining path, the unmanned vessel control terminal will automatically trigger an alarm mechanism, send alarm information to the background monitoring platform, and activate the emergency plan to avoid navigation risks or complete mission interruption for the unmanned vessel.

[0131] Furthermore, after completing a single monitoring task, the dynamic cost field is updated according to the local navigation log and adaptive adjustment record, so that the dynamic cost field can match the actual water environment resistance changes in the target water area in real time, avoiding energy consumption estimation deviations or navigation risks in subsequent path planning due to reliance on lagging resistance data.

[0132] In the method provided in this application embodiment, "updating the dynamic cost field according to the local navigation log and adaptive adjustment record" includes:

[0133] Acquire absolute displacement from GPS, attitude from IMU, and thrust and torque from propulsion motors;

[0134] Estimate the inversion velocity and inversion direction on the global optimization path and store them in an inversion database;

[0135] Based on the preset update cycle constraint, the relative resistance coefficient distribution of the dynamic cost field is calculated by combining the inversion database update.

[0136] In this embodiment of the application, in order to make the dynamic cost field fit the actual water environment resistance changes of the target water area in real time, it is necessary to extract key data during the navigation process to invert the water flow velocity and direction, and then update the dynamic cost field based on the inversion database, so as to improve the accuracy of subsequent global path optimization and the economy of unmanned vessel navigation.

[0137] Specifically, the system first obtains the absolute displacement provided by GPS, the attitude provided by IMU, and the thrust and torque provided by the propulsion motor.

[0138] Among them, GPS records the changes in latitude and longitude coordinates of the unmanned ship in real time during the navigation process, forming a continuous absolute displacement trajectory to intuitively reflect the position deviation of the ship's actual navigation; IMU (Inertial Measurement Unit) collects the ship's attitude information, including heading angle, pitch angle, roll angle, etc., and then determines whether the ship's attitude has deviated due to water flow impact or wind and waves.

[0139] Furthermore, the thrust and torque data of the propulsion motors are directly related to the ship's power output, and changes in power output can indirectly reflect fluctuations in water resistance. For example, when the resistance in a certain section of water increases, the propulsion motors need to increase thrust and torque to maintain the preset sailing speed. The above three types of data are stored in real time in the unmanned vessel's local navigation log, providing basic data support for subsequent inversion calculations.

[0140] Furthermore, the inversion velocity and inversion direction on the global optimization path are estimated and stored as an inversion database. That is, the above three types of data are fused by the state estimation algorithm in the existing technology.

[0141] Specifically, based on the absolute displacement trajectory of GPS, the actual navigation direction of the ship is corrected by combining the attitude data of IMU to eliminate the interference of attitude deviation on the displacement data. Then, the corrected displacement data is combined with the thrust and torque data of the propulsion motor, and the water flow velocity and direction acting on the ship are calculated in reverse according to the principles of fluid mechanics and the hull parameters of the unmanned ship. These are used as the inversion velocity and inversion direction.

[0142] For example, when the actual speed of the ship is lower than the preset speed and the thrust of the propulsion motor has been increased to the preset thrust threshold, the magnitude of the reverse flow velocity of the water flow can be deduced by calculating the difference between the actual speed of the ship and the preset speed, the difference between the actual thrust of the propulsion motor and the preset thrust threshold, and combining the ship drag coefficient and the hydrodynamic model.

[0143] In addition, when the hull experiences lateral displacement deviation, the direction of the lateral water flow can be deduced by combining the hull attitude data such as heading angle and roll angle collected by the IMU, as well as the thrust distribution data of the propulsion motor in the lateral direction, through force balance analysis.

[0144] Meanwhile, the flow velocity and flow direction data obtained from each monitoring task are classified and stored according to the monitoring point and timestamp to form an inversion database, so as to ensure the traceability and reusability of the data.

[0145] Furthermore, based on a preset update cycle constraint, the distribution of the relative drag coefficient of the dynamic cost field is calculated by updating the inversion database. The preset update cycle needs to be determined in conjunction with the frequency of water environment changes in the target water area. For example, for lakes with gentle water flow changes, the dynamic cost field can be updated once every 5 monitoring tasks; for rivers or estuaries with frequent water flow changes, it can be updated once every 2 monitoring tasks or every 12 hours, to ensure that the dynamic cost field can reflect water area changes in a timely manner.

[0146] Specifically, firstly, all inversion velocity and direction data covered within the update cycle are extracted from the inversion database, and then data matching is performed at points with preset intervals in the dynamic cost field to find the inversion data of each point in different navigation directions.

[0147] Furthermore, using the baseline navigation resistance in static water as a reference, the relative resistance coefficient of each point in different directions is recalculated based on the inverted flow velocity and direction data. For example, if the inverted flow velocity at a certain point is 0.5 m / s and the flow direction is opposite to the preset navigation direction, the relative resistance coefficient in that direction needs to be adjusted upward based on the baseline value according to the preset formula.

[0148] Conversely, if the inverted flow direction is the same as the preset navigation direction, the relative drag coefficient for that direction is appropriately lowered. The recalculated relative drag coefficient distribution replaces the corresponding data in the original dynamic cost field, completing the cost field update so that the updated cost field can more accurately reflect the actual resistance of the current water area.

[0149] Finally, after completing a single monitoring task, the dynamic cost field is updated according to the local navigation log and adaptive adjustment record to ensure that when the unmanned vessel conducts path planning in the future, it can formulate an optimized path based on the latest water environment resistance data, reduce energy waste or navigation risks caused by the lag of the cost field, and provide a more reliable environmental parameter reference for the adaptive adjustment mechanism, thereby improving the overall monitoring efficiency and navigation stability of the unmanned vessel.

[0150] The embodiments of this application, through the specific implementation methods described above, achieve the following technical effects:

[0151] This application proposes an autonomous navigation path planning method for unmanned surface vessels (USVs) used for water quality monitoring. First, based on historical prior knowledge of each monitoring point within the target water area, the water quality fluctuation coefficient is calculated and weighted with location importance to obtain a priority score. This score is then combined with a pre-set monitoring frequency mapping table to determine the monitoring frequency for each point. Next, the least common multiple of the monitoring frequencies is calculated to set the list cycle. Task time slices are divided based on the USV's endurance characteristics, and sampling timestamps are distributed according to the monitoring frequency. The sampling points in each time slice are iterated to form a monitoring list cycle. Then, in response to the monitoring task start command, a real-time monitoring list containing time matching information is extracted from the monitoring list cycle, verified, and adapted before being sent to the USV. Subsequently, historical flow velocity and direction data of the target water area are acquired to generate a flow velocity-direction distribution map. The relative drag coefficient distribution is calculated to construct an anisotropic dynamic cost field. The feasibility of historical paths is verified; if feasible, they are reused; otherwise, the dynamic cost field is used as the energy consumption basis, and point priority as the benefit metric. An endurance constraint and a greedy algorithm are combined to solve for a globally optimized path. Finally, the path is issued to control the unmanned vessel to collect water quality data, and the path is adjusted adaptively in real time. The flow velocity and direction are inverted and stored in an inversion database. The dynamic cost field is updated according to a preset period to provide accurate environmental parameters for subsequent planning.

[0152] The method provided in this application, through the technical solution of "determining the monitoring frequency and inventory cycle - extracting the real-time monitoring inventory - constructing a dynamic cost field and path optimization - performing navigation adjustment and cost field update", solves the problems of traditional ship path planning, such as difficulty in accurately considering the complex changes in the water environment, the disconnect between the path and actual navigation needs, resulting in high energy consumption, low efficiency, and lack of effective adaptive adjustment mechanism. It provides technical support for long-term, stable and efficient water quality monitoring of target waters.

[0153] Example 2, as shown in the appendix Figure 2 As shown, based on the inventive concept of the autonomous navigation water quality detection unmanned surface vessel path planning method provided in Embodiment 1, this application also provides an autonomous navigation water quality detection unmanned surface vessel path planning system, specifically including:

[0154] The monitoring frequency and list generation module 01 is used to determine the monitoring frequency of each monitoring point based on the historical prior knowledge of each monitoring point in the target water area, and generate a monitoring list cyclically according to the monitoring frequency.

[0155] The real-time monitoring list extraction module 02 is used to respond to the monitoring task start command and extract the real-time monitoring list that needs to be executed from the monitoring list in a loop.

[0156] The cost field and path optimization module 03 is used to obtain a dynamic cost field that characterizes the water environment resistance in the target water area, and to perform global path optimization based on a multi-objective optimization method, combined with the real-time monitoring list and the dynamic cost field, to obtain a globally optimized path.

[0157] The path distribution navigation module 04 is used to distribute the globally optimized path to the unmanned vessel for local navigation and adaptive adjustment.

[0158] In one embodiment, the monitoring frequency and inventory generation module 01 is further configured to:

[0159] The process involves acquiring historical prior knowledge, including historical water quality monitoring data and location importance weights for the target area; calculating water quality fluctuation coefficients based on the historical water quality monitoring data, and weighting and fusing the water quality fluctuation coefficients with the location importance weights to obtain a priority score for each monitoring point; determining the numerical range to which the priority score belongs by combining a preset monitoring frequency mapping table, and identifying the monitoring frequency level and corresponding monitoring frequency for each monitoring point; calculating the least common multiple of the monitoring frequencies of multiple monitoring points, and using the duration corresponding to the least common multiple as the list cycle period; dividing the list cycle period into multiple continuous and fixed-duration task time slices based on the typical endurance characteristics of unmanned surface vessels; distributing sampling timestamps according to the monitoring frequencies of multiple monitoring points within the list cycle period; for each task time slice, identifying sampling objects by traversing multiple monitoring points based on the sampling timestamp distribution results, and adding the monitoring points to the monitoring list corresponding to each task time slice to form the monitoring list cycle.

[0160] In one embodiment, the cost field and path optimization module 03 is also used for:

[0161] Historical flow velocity and direction data of the target water area are acquired to generate a flow velocity-direction distribution map. Using the baseline navigation resistance in static water as a reference, the relative resistance coefficient distribution of multiple points at equal preset intervals within the target water area is calculated based on the flow velocity-direction distribution map, forming the dynamic cost field. The relative resistance coefficient distribution includes multiple relative resistance coefficients in multiple navigation directions with equal preset azimuth angle differences, and the relative resistance coefficient distribution at each point exhibits anisotropy. The dynamic cost field serves as the basis for path energy consumption calculation. The priority score corresponding to each monitoring point in the real-time monitoring list is used as the information collection benefit metric. An optimization function is constructed with the objectives of minimizing total path energy consumption and maximizing path energy efficiency and path information collection efficiency. This function, combined with unmanned surface vessel endurance constraints and a path optimization algorithm, solves for the optimal path sequence to access the real-time monitoring list, outputting the globally optimized path.

[0162] Furthermore, the cost field and path optimization module 03 also includes:

[0163] Retrieve the historical path scheme corresponding to the real-time monitoring list from the previous monitoring list loop; combine the dynamic cost field to verify the feasibility of the historical path scheme, including: verifying the geographical feasibility of the historical path scheme, and determining whether the maximum point of the relative drag coefficient of the historical path scheme under the dynamic cost field meets the preset maximum drag coefficient threshold; if the geographical feasibility verification is successful, further verify the endurance feasibility of the historical path scheme, estimate the total energy consumption of sailing along the historical path scheme based on the dynamic cost field, and determine whether the total energy consumption is within the preset safe endurance threshold; if both geographical feasibility and endurance feasibility are verified to be successful, then the historical path scheme is adopted as the global optimized path; if either geographical feasibility or endurance feasibility fails verification, then global path optimization is performed.

[0164] In one embodiment, the path distribution navigation module 04 is also used for:

[0165] The unmanned surface vessel is controlled to navigate along the globally optimized path and collect water quality data. During the navigation, the globally optimized path is adaptively adjusted in real time based on the energy consumption rate and the local endurance estimation model. The dynamic cost field is updated accordingly based on the local navigation log and the adaptive adjustment record.

[0166] Furthermore, the path distribution navigation module 04 also includes:

[0167] The system acquires the absolute displacement provided by GPS, the attitude provided by IMU, and the thrust and torque provided by the propulsion motor; it estimates the inversion velocity and inversion direction on the global optimization path and stores them in an inversion database.

[0168] Based on the preset update cycle constraint, the relative resistance coefficient distribution of the dynamic cost field is calculated by combining the inversion database update.

[0169] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0170] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0171] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. An autonomous cruise water quality detection unmanned ship path planning method, characterized in that, The method comprises the following steps: determining the monitoring frequency of each monitoring point based on the historical prior knowledge of each monitoring point in the target water area, and generating a monitoring list cycle according to the monitoring frequency; in response to a monitoring task start instruction, extracting a real-time monitoring list currently to be executed from the monitoring list cycle; obtaining a dynamic cost field representing the water environment resistance in the target water area, and performing global path optimization based on a multi-objective optimization method, combining the real-time monitoring list and the dynamic cost field to obtain a globally optimized path; downloading the globally optimized path to the unmanned ship to perform local navigation and adaptive adjustment; wherein, based on the historical prior knowledge of each monitoring point in the target water area, the monitoring frequency of each monitoring point is determined, comprising: obtaining the historical prior knowledge, wherein the historical prior knowledge includes historical water quality monitoring data and site importance weight of the target area; calculating the water quality fluctuation coefficient according to the historical water quality monitoring data, and weighting and fusing the water quality fluctuation coefficient and the site importance weight to obtain the priority score of each monitoring point; combining a preset monitoring frequency mapping table to determine the numerical interval to which the priority score belongs, and determining the monitoring frequency level and the corresponding monitoring frequency of each monitoring point; wherein, combining the real-time monitoring list and the dynamic cost field to perform global path optimization and obtain a globally optimized path, comprising: taking the dynamic cost field as the path energy consumption calculation basis; taking the priority score of each monitoring point in the real-time monitoring list as the information collection benefit measure; constructing an optimization function with the goal of minimizing the total path energy consumption, maximizing the path energy consumption efficiency and the path information collection efficiency, combining the unmanned ship endurance constraint and the path optimization algorithm to solve the optimal path sequence for visiting the real-time monitoring list, and outputting the globally optimized path.

2. The autonomous cruise water quality detection unmanned ship path planning method of claim 1, wherein, According to the monitoring frequency, a monitoring list cycle is generated, comprising: calculating the numerical least common multiple of the monitoring frequencies of multiple monitoring points, and taking the time length corresponding to the numerical least common multiple as the list cycle period; combining the typical endurance characteristics of the unmanned ship, dividing the list cycle period into multiple continuous and fixed-length task time slices; according to the monitoring frequencies of multiple monitoring points, distributing the sampling time stamps in the list cycle period; for each task time slice, according to the sampling time stamp distribution result, traversing multiple monitoring points to determine the sampling object, and adding the monitoring point to the monitoring list corresponding to each task time slice to form the monitoring list cycle.

3. The path planning method for autonomous cruise water quality detection unmanned ship according to claim 2, characterized in that, Obtaining a dynamic cost field representing the water environment resistance in the target water area, comprising: obtaining historical flow rate and flow direction data of the target water area to generate a flow rate-flow direction distribution map of the target water area; taking the reference navigation resistance in a static water body as a reference, calculating the relative resistance coefficient distribution of multiple points with a preset interval in the target water area according to the flow rate-flow direction distribution map to form the dynamic cost field; wherein, the relative resistance coefficient distribution includes multiple relative resistance coefficients in multiple navigation directions with a preset azimuth angle difference, and the relative resistance coefficient distribution of each point has anisotropy.

4. The path planning method for autonomous cruise water quality detection unmanned ship according to claim 1, characterized in that, acquiring a dynamic cost field representing water environment resistance in the target water area, and then comprising: calling a historical path scheme corresponding to the real-time monitoring list in the last monitoring list cycle; verifying the feasibility of the historical path scheme in combination with the dynamic cost field, comprising: verifying the geographical feasibility of the historical path scheme, judging whether a relative resistance coefficient maximum point of the historical path scheme under the dynamic cost field meets a preset maximum resistance coefficient threshold; if the geographical feasibility verification passes, further verifying the endurance feasibility of the historical path scheme, estimating total energy consumption of sailing along the historical path scheme based on the dynamic cost field, and judging whether the total energy consumption is within a preset safe endurance threshold; if both the geographical feasibility and the endurance feasibility pass, adopting the historical path scheme as the global optimization path; if either of the geographical feasibility and the endurance feasibility fails, performing global path optimization.

5. The path planning method for autonomous cruise water quality detection unmanned ship according to claim 1, wherein, issuing the global optimization path to the unmanned ship to perform local navigation and adaptive adjustment, further comprising: controlling the unmanned ship to sail along the global optimization path and collect water quality data; in the sailing process, performing real-time adaptive adjustment on the global optimization path based on an energy consumption rate and a local endurance estimation model; correspondingly updating the dynamic cost field according to a local navigation log and an adaptive adjustment record.

6. The autonomous cruise water quality detection unmanned ship path planning method of claim 5, wherein, correspondingly updating the dynamic cost field according to a local navigation log and an adaptive adjustment record, comprising: acquiring absolute displacement provided by GPS, attitude provided by IMU, thrust and torque provided by a propulsion motor; estimating inversion flow velocity and inversion flow direction on the global optimization path, and storing as an inversion database; based on a preset update period constraint, updating a relative resistance coefficient distribution of the dynamic cost field in combination with the inversion database.

7. An autonomous cruise water quality detection unmanned ship path planning system, characterized in that, The system is used to perform the autonomous cruising water quality detection unmanned ship path planning method of any one of claims 1-6, and the system comprises: a monitoring frequency and list generation module, configured to determine a monitoring frequency of each monitoring point based on historical prior knowledge of each monitoring point in a target water area, and generate a monitoring list cycle according to the monitoring frequency; a real-time monitoring list extraction module, configured to extract a real-time monitoring list to be executed currently from the monitoring list cycle in response to a monitoring task start instruction; a cost field and path optimization module, configured to acquire a dynamic cost field representing water environment resistance in the target water area, and perform global path optimization based on a multi-objective optimization method in combination with the real-time monitoring list and the dynamic cost field, to acquire a global optimization path; a path issuing and navigation module, configured to issue the global optimization path to the unmanned ship to perform local navigation and adaptive adjustment.

Images