China 
Africa 
Explore chapters and articles related to this topic
Introduction to Expert Systems
Published in Chris Nikolopoulos, Expert Systems, 1997
Another major problem, which hinders the effectiveness of logic programming systems, is finding the appropriate search strategy to shorten the proof of a theorem. An exhaustive (brute force) search can be prohibitive in both time and space considerations. A search strategy needs to be employed which determines the clauses that are to be chosen for resolution, among the many possible resolvable pairs of clauses. A search strategy can make a very important contribution to the overall effectiveness of the system. A good search strategy significantly shortens the time needed to reach a proof, by not involving resolutions of clauses irrelevant to the proof. We give below some examples of search strategies which have been proposed in the literature:
Multi-Objective Optimization Algorithms for 72 Pin Fin Heat Sinks
Published in Srikanth Rangarajan, C. Balaji, Phase Change Material-Based Heat Sinks: A Multi Objective Perspective, 2019
Srikanth Rangarajan, C. Balaji
From this study, it is evident that the brute-force search method gives a picture of the objective function space and hence a clear idea of the true Pareto optimal solution. The major drawback of the brute force search method is the large number of points required to generate the objective function space and high computational time. However, this method can be used to test any multi-objective algorithm, in terms of its closeness to the true Pareto optimal solution. Table 6.4 reports a critical comparison of the performance of the four optimization algorithms implemented in this study.
Real-Time Active SLAM and Obstacle Avoidance for an Autonomous Robot Based on Stereo Vision
Published in Cybernetics and Systems, 2019
V. S. Kalogeiton, K. Ioannidis, G. Ch. Sirakoulis, E. B. Kosmatopoulos
The Grid-based exhaustive search (GBES) algorithm was used to calculate the theoretically optimal solution. The visual space with the stereo camera is discretized and a search is performed over all possible combinations so that the objective function J is minimized. Ideally, GBES should return the global optimal solution and so, it could be used as a benchmark for evaluating the performance of every algorithm only if the grid size is sufficiently small. In practice, nonetheless, it is difficult to guarantee a solution since its computational complexity and memory requirements are rather exponential in the number of under study environment size. Hence, implementing the GBES becomes prohibitive, almost non feasible, when the under study environment has an increased size and/or when the size of the grid cells is rather small. In general, brute-force search consists of systematically enumerating all candidates for the solution and checking whether each candidate satisfies the problem’s statement. Therefore, its cost is proportional to the number of candidate solutions and so it is usually exploited only when the problem size is limited.
Vision-based surface roughness evaluation system for end milling
Published in International Journal of Computer Integrated Manufacturing, 2018
Besmir Cuka, Minho Cho, Dong-Won Kim
It should be noted that all the steps described above need to be applied only for the largest detected contour on the first iteration. For the remaining contours, the radius of the first fitted circle should be maintained and only the centre of the new circle should be estimated. The radius is only estimated on the first iteration because the largest detected contour is more likely to represent the shape of a tool mark, whereas the smaller contours need to be combined to provide a reliable result. In the case where only the centre of the circle needs to be determined, two approaches can be considered. The first approach involves using a brute force search, where each pixel of the image is considered as the possible centre of the circle. Thus, for each point on the image, a circle is drawn with the centre of the coordinates of the point, and the fitness of the circle on the targeted contour is then evaluated. The brute force search ensures an optimal solution but is computationally intense and time consuming. The second approach involves using Equations (8)–(10), with radius being a known parameter. Examples of detected tool marks from two different surfaces are shown in Figure 5. Both images are filtered and thresholded by using the same parameters and most of the tool marks are successfully detected even though their orientation, size and thickness significantly differed. The density of the detected contours (tool marks) and their distribution in relation to the feed per tooth can determine the cutting condition of the tool at the moment of the machining operation. Furthermore, the orientation of the texture, the size of the tool or the scale of the image does not affect the final evaluation of the inspected surface.