Research On Measurement Sequence Planning Of Free-form Surface Based On Firefly Algorithm

Free-form surface measurement techniques occupy an increasingly importantposition in the field of modern machinery manufacturing. And it is the research focusin the field of precision measurements in recent years. Since the coordinate measuringmachine with high precision, most of the free-form surface measurements arecompleted by CMM-contact measurement in the current industrial applications.However, because of inefficiency and low intelligence, CMM is restricted in the largescale free-form surface measurement tasks. To solve the problem, it is one of theeffective methods where facilities which are highly intelligent, independent foraccurate CAD and prior knowledge are equipped to CMM.For autonomous measurement sequence planning problem of free-form surface,objective function considering the efficiency, rationality and economy is proposed inthis paper. Then firefly algorithm which is swarm-based is applied to solve theproblem: Firstly, encoding rules of the standard firefly algorithm, location update andthe iterative rules are improved to solve the measurement sequence planning problem.In addition, for multi-measurement terminal assignment problem, the discrete fireflyalgorithm is improved further.In this paper, the work done is as follows:1. A free-form surface measurement sequence planning mathematical model isestablished. On the basis of the shortest measuring path length, combining with thefeatures of CMM measurement process for freeform surface, the objective functionconsidering the smoothness of the measurement path and the probe rotation distanceis established.2. Improved firefly algorithm is proposed and applied to freeform surfacemeasurement sequence planning problem. To solve the problem of free-form surfacemeasurement sequence planning, discrete operations are conducted and a newdefinition of distance mode is proposed. Also the location update mechanism of thealgorithm is redefined and moving rules are improved to enhance the global searchability. So that, standard firefly algorithm is improved to be discrete firefly algorithm.The general methods and steps are presented to solve measurement sequence planningand measuring sequence planning trials with two free surface parts are carried out. The results show that modified discrete firefly algorithm applies to the measuringsequence planning problem of freeform surface. The contrast test further demonstratessignificantly better than the optimization ability of genetic algorithms.3. A mathematical model about the multi-measure terminal assignment problem isestablished. The discrete firefly algorithm is improved further to adapt to the problem.Under the average distribution of tasks, a mathematical model of the measurementtask allocation problem of multi-measurement terminal is established in free-formsurface digitization process. Then, the discrete firefly algorithm encoding rules,distance definition and moving rules are improves to adapt to such problem. Finally,taking hk48problem and another free surface part for task allocation test,experimental results show that the algorithm is suitable for solving such taskallocation problem.