Two-step procedure of optimisation for flight planning problem for airborne LiDAR data acquisition

Airborne LiDAR with simultaneous photographic data acquisition is an effective technique for 3D data collection with colour or textural information. Flight planning for airborne LiDAR data collection ensures the desired qualities in the data. While the desired quality of data is enforced by adopting the appropriate flight planning parameters during the flight operations, attempt is also made to minimize the flight duration which leads to minimizing the cost of the project. The flight duration is expressed as the sum of strip and turning time. The data requirements, preferences and limitations, which are associated with a flight planning problem, are framed as the constraints. The scanner parameters (half scan angle, scanning frequency, PRF) and flying parameters (flying height, flying speed, flying direction) are identified as variables of the optimization. Due to the typical characteristics of flight duration and scanner parameters, and the absence of the initial solution, the classical methods of optimization are found not deployable. Consequently, genetic algorithms (GA), which is an evolutionary algorithm, is employed as an alternative procedure for solving the problem. However, GA usually demands longer time for the convergence. Therefore, considering the pitfalls of both GA and classical method, a two-step procedure, which consists of GA and Hooke and Jeeve’s (HJ) classical method of optimization, is suggested. The two-step procedure is implemented for a simulated AOI (Area of Interest for data capture) and an AOI of an actual test site with different data requirements. Results obtained by the GA show higher confidence in the scanner parameters which are mathematically discrete. Considering the scanner parameters as constant in the HJ method, the results of GA are