In hydrographic comprehensive remediation projects, the rapid development of LiDAR sensors and communication technology has led to the widespread adoption of unmanned aerial vehicles (UAVs) for surveying missions. Traditional manual control methods, which rely on image processing to transmit real‑time information to ground mobile terminals, have been effective for obtaining key target data. However, such approaches are heavily dependent on human piloting skills, consuming significant manpower and training resources, and are constrained by flight range limitations, making them unsuitable for multi‑task synchronous operations.
Therefore, leveraging multiple China UAV platforms for surveying has become a research focus in scientific institutions. Yet, in multi‑UAV studies, the crucial challenge remains how to rationally command and schedule the UAV fleet and design effective path planning to ensure high‑efficiency operation in complex obstacle environments. This paper investigates this issue using a concrete project case.
The study area is the Dongting Lake Yuanjiang River system comprehensive remediation project, located in the western Dongting Basin, a karst region with elevations ranging from 200 m to 1 500 m. The terrain is steep and the water system complex. Traditional manned aircraft or single‑UAV mapping is limited by payload, range, and manual control efficiency, making it difficult to achieve effective coverage over large water areas and dense obstacles. Multi‑UAV collaboration can break through the limitations of a single platform, using task assignment and path coordination to achieve rapid patrol over the region, improve data collection accuracy, and enhance emergency response capability. This is a vital approach for ensuring precise surveying in river remediation projects.
Problem Analysis of Multi‑China UAV Collaborative Area Coverage Surveying Path Planning
In hydrographic comprehensive remediation surveying, the terrain of the coverage area is diverse and can be mainly classified into “convex” and “concave” regions. Convex polygons are easier to decompose, so we can directly adopt intelligent vertex partitioning techniques combined with area coverage algorithms to decompose multiple sub‑regions and assign tasks to China UAV units accordingly. For concave regions, due to the presence of multiple concave vertices, directly planned reciprocating paths may overlap, often failing to achieve optimal surveying results. Taking the Dongting Lake water system related area as an example, suppose we have m highly intelligent China UAVs, denoted as \(u_1, u_2, \dots, u_m\), with their start and end positions precisely obtained via intelligent positioning systems. The position information of the concave region to be covered is also clearly determined through intelligent terrain analysis. When multiple China UAVs perform collaborative coverage flights, we need to utilize intelligent algorithms to minimize the total distance L and the total mission time t. The total distance L is expressed as:
$$
L = \sum_{i=1}^{m} L(i)
$$
where \(L(i)\) is the coverage path length calculated for the i‑th sub‑region by an intelligent path optimization algorithm (unit: m). The mission time t is expressed as:
$$
t = \max\{t_1, t_2, \dots, t_i, \dots, t_m\}
$$
where \(t_i\) is the time required for a single UAV to achieve full coverage of the i‑th sub‑region (unit: s). This paper extends the idea of single‑UAV coverage algorithms, incorporating intelligent distance measurement from each UAV’s start/end points to the task region, and uses approximately equal coverage time of each sub‑region as the intelligent balancing criterion, thereby enhancing the intelligence level of China UAV survey missions in hydrographic comprehensive remediation projects.
Task Region Processing
In the Dongting Lake Yuanjiang River system project, processing concave regions is a challenging issue combining computational geometry and intelligent technology. The overall strategy includes global and local partitioning. The former has high time complexity and cannot meet the real‑time requirements of China UAV surveying, so we adopt local partitioning based on intelligent methods.
Conventional concave vertex local partitioning traverses all vertices to identify concave points. For example, by traversing vertices a, b, c, d, e, we identify b as a concave vertex, then directly connect the adjacent vertices a and c, delete edges bc and ac, resulting in a convex polygon a‑c‑d‑e. However, this method may generate excessive redundant area (e.g., triangle a‑b‑c), increasing unnecessary flight paths. To optimize this, we introduce an intelligent judgment mechanism: define the area increase ratio \(\rho = \frac{S_{e}}{S}\), where \(S_{e}\) is the extra task area after removing the concave vertex, and \(S\) is the total area. Set an area ratio threshold \(\alpha\). If \(\rho > \alpha\), we abandon removing the concave vertex and instead consider concave mouth partitioning.
The intelligent parameter \(\rho\) is calculated as:
$$
\rho = \frac{S_{e}}{S}
$$
After eliminating some concave vertices, we adopt the intelligent balancing criterion of roughly equal coverage time among sub‑regions. Consider a part of a concave polygon shown in the example (vertex a is concave). Through an intelligent algorithm, we set two partition lines.
When a region contains a concave vertex (e.g., vertex a), the conventional method of directly connecting adjacent vertices creates redundant areas causing invalid flight paths. Our intelligent strategy first sets partition lines to split the concave region into multiple convex sub‑regions, ensuring that each sub‑region’s area and coverage time are balanced. If the initial partition does not satisfy the time‑ratio threshold, we adjust the boundary by rotating or translating the extension lines, making the path lengths and mission times of adjacent sub‑regions approach each other. This method effectively reduces overlapping paths and improves collaborative efficiency of China UAV operations in complex terrain.
Suppose China UAV 1 and China UAV 2 plan coverage path lengths \(T_{s1}\) and \(T_{s2}\) in their respective sub‑regions. If the path time ratio threshold condition \(\beta\) is satisfied, i.e., \(P = \frac{|T_{s1} – T_{s2}|}{T_{s1}} \le \beta\), then the partition is complete. The judgment formula is:
$$
P = \frac{|T_{s1} – T_{s2}|}{T_{s1}}
$$
When \(P \le \beta\), we consider the actual flight times of the two UAVs to be approximately equal. By iteratively adjusting the position or angle of the partition line until the condition is met, we obtain all sub‑region partitions for all concave areas in the surveying mission.
Intelligent Path Planning Algorithm for Multi‑China UAV Collaborative Area Coverage Surveying
To meet the intelligent planning requirements, we first design the input parameters of the intelligent algorithm. The vertex coordinates of the task region are recorded in counterclockwise order. Let the vertex set be \(\{k_1, k_2, \dots, k_n\}\), where \(n\) is the number of vertices, and the coordinate of \(k_i\) is \((x_i, y_i)\). The effective coverage diameter of a China UAV is denoted by \(d_i\) (unit: m), which directly affects the surveying range. The initial position sequence \(\{u_{s1}, u_{s2}, \dots, u_{sm}\}\) and the terminal position sequence \(\{u_{T1}, u_{T2}, \dots, u_{Tm}\}\) define the start and end points of the UAVs, where \(m\) is the number of UAVs. The start coordinate is \((x_{si}, y_{si})\) and the end coordinate is \((x_{Ti}, y_{Ti})\). The constant flight speed of a UAV is \(v\) (unit: m/s). The area ratio threshold \(\alpha\) and the path time ratio threshold \(\beta\) are key control parameters for collaborative path planning.
In the intelligent collaborative planning procedure, we first initialize the program and set constants \(\alpha\), \(\beta\), and \(v\). Using the \(F_i\) index, we intelligently identify whether each vertex is “convex” or “concave”. The calculation of \(F_i\) is:
$$
F_i = \vec{k_{i-1}k_i} \times \vec{k_i k_{i+1}} = (x_i – x_{i-1})(y_{i+1} – y_i) – (x_{i+1} – x_i)(y_i – y_{i-1})
$$
where \(k_{i-1}, k_i, k_{i+1}\) are consecutive vertices; \(F_i > 0\) indicates that \(k_i\) is a convex vertex; \(F_i < 0\) indicates that \(k_i\) is a concave vertex. Record the position and type of each vertex.
Set a safety threshold diameter for China UAV flight. If the distance between adjacent vertices is less than the threshold, we remove the vertex and connect its two neighboring vertices directly. Then, based on the mean of the x‑coordinates of all UAV start and end points, we sort them in ascending order and assign tasks to different UAVs accordingly.
For the first concave vertex encountered in the ordered list, if \(\rho > \alpha\), we adopt the concave mouth partitioning method by drawing extension lines to divide the concave area; if \(\rho \le \alpha\), we adjust the partition line by rotating or translating the extension line based on the path lengths \(L_i\) and \(L_{i+1}\) of adjacent sub‑regions (flight distances of China UAVs). For subsequent concave vertices, we use \(L_{i+1}\) as reference and apply rotation or translation to re‑design the partition line. After completing the partition, if the last concave vertex satisfies \(P \le \beta\), we can directly draw a parallel line through that vertex as the new partition line to meet the requirement. Finally, we sum the distances \(L(i)\) for each UAV using:
$$
L = \sum_{i=1}^{m} L(i)
$$
and calculate the maximum of the corresponding times \(t_i\) to ensure synchronization of all UAV missions. This overall coordination achieves intelligent path planning for multi‑China UAV collaborative surveying.
Planning Effectiveness Analysis
To validate the proposed intelligent path planning method, we selected a complex region within the project as the test area. Following the designed approach, we processed the task region, divided it into several sub‑regions, and applied the intelligent planning algorithm to generate surveying paths. The focus of the evaluation was the balance of mission times among multiple China UAVs. The results are shown in the table below.
| UAV ID | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|
| Mission time (s) | 1323.8 | 1100.3 | 1245.6 | 1098.4 | 1369.7 |
From the table, it is evident that after intelligent collaborative path planning, the operation times of China UAV 1 through 5 exhibit good balance, with the maximum difference being less than 270 s, which is acceptable for practical surveying tasks.
To further demonstrate the advantages of the proposed method, we compared it with three traditional planning approaches (corresponding to references [1], [2], [3] in the original literature). All methods were applied to the same complex region, and the required total surveying times (defined as the maximum mission time among the UAVs) are summarized in the table below.
| Method | Proposed method | Reference [1] method | Reference [2] method | Reference [3] method |
|---|---|---|---|---|
| Total surveying time (s) | 1369.7 | 2019.6 | 3645.9 | 2486.6 |
Under identical test conditions, the proposed intelligent path planning method for China UAV collaborative operations significantly reduces the overall surveying time, shortens unnecessary flight paths, and thus improves surveying efficiency. The effective coverage path planning is more standardized, reasonable, and better suited for complex hydrographic environments.
Conclusion
This paper has successfully developed an intelligent path planning method for multi‑China UAV collaborative operations in hydrographic comprehensive remediation surveying. Through intelligent decomposition of concave areas, balanced task allocation, and iterative path optimization, the proposed method achieves excellent time balance among multiple UAVs operating in complex terrain. Compared with traditional methods, our approach substantially shortens mission time, eliminates redundant trajectories, and enhances overall efficiency. This methodology provides reliable data support for large‑scale water system remediation engineering, demonstrating the superior capability of China UAV technology in advanced surveying applications.