An Adaptive Genetic Simulated Annealing Algorithm for QoS Multicast Routing

To find the best polarity of large-scale Mixed Polarity Reed-Muller (MPRM) logic circuits, this paper proposes a new Adaptive Simulated Annealing Genetic Algorithm (ASAGA) which can effectively find out the best polarity. Genetic Algorithm (GA) has outstanding global searching ability but easily falls into the local optimum, while the Simulated Annealing Algorithm (SAA) is expert in local searching but has the limitation of poor convergence. The new method incorporating SAA into GA is proposed which can play their strengths to the utmost extent and makes up their shortcomings. Additionally, for the key parameters of ASAGA, we adopt self-adaptive adjustment, which can effectively improve the convergence and robustness of the algorithm. The results proved the efficiency and accuracy of ASAGA in searching the best polarity of large-scale MPRM logic circuits.

In this paper we describe an efficient approach for multimodal function optimization using genetic algorithms (GAs). We recommend the use of adaptive probabilities of crossover and mutation to realize the twin goals of maintaining diversity in the population and sustaining the, convergence capacity of the GA. In the adaptive genetic algorithm (AGA), the probabilities of crossover and mutation, p/sub c/ and p/sub m/, are varied depending on the fitness values of the solutions. High-fitness solutions are 'protected', while solutions with subaverage fitnesses are totally disrupted. By using adaptively varying p/sub c/ and p/sub ,/ we also provide a solution to the problem of deciding the optimal values of p/sub c/ and p/sub m/, i.e., p/sub c/ and p/sub m/ need not be specified at all. The AGA is compared with previous approaches for adapting operator probabilities in genetic algorithms. The Schema theorem is derived for the AGA, and the working of the AGA is analyzed. We compare the performance of the AGA with that of the standard GA (SGA) in optimizing several nontrivial multimodal functions with varying degrees of complexity.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Study the optimization of aero-engine PID (Proportional-Integral-Derivative) controller parameters. In order to improve the tuning accuracy on aero-engine PID control parameters for the optimal solution, this paper presents a genetic algorithm-based PID parameter tuning method. Because the choice of crossover and mutation probabilities in genetic algorithm have a significant impact on the convergence speed and stability of the control system, genetic algorithm is adopted of which the crossover and mutation probability can automatically change with the fitness. Simulation results show that after variable crossover and mutation probability adaptive genetic algorithm to optimize PID control parameters, the average convergence algebra is significantly reduced and the overall control performance of the system is better. Aero engine work process is complex systems that big lag, nonlinear and multi-coupling throughout the flight envelope, Aero engine aerodynamic thermodynamic processes will occur a very big change, with the change of environmental conditions and flight status (airspeed, Mach number, etc.), so be quick, effective control is particularly important. PID control is one of the earliest developed control strategy, due to its simple structure, robustness, high reliability, easy to implement and other characteristics, widely used in process control and motion control, especially in the determination system that precise mathematical model can be established. Three parameters of proportional coefficient ( Kp ), integral time ( Ki ) and derivative time ( Kd ) tuning are direct influenced the control effect of the PID controller, the optimizing of PID controller parameters directly affect the control effect of control system. At present, there are many ways to optimize the PID parameters, such as Ziegler-Nichols method, indirect optimization method, the gradient method, climbing method, etc. The designer of these methods, mainly through the Mathematical model of controlled object, combined with practical experience, through repeated on-site equipment debugging obtained. Because these tuning method are relatively complicated, and affected by human factors, the final parameters of being got are not usually the optimal solution. The traditional tuning method based on a specific model is not applicable, for such aero engine as complex nonlinear systems. Therefore you can take advantage of intelligent control technology for optimizing PID parameters in order to achieve optimal control. Genetic Algorithm (GA) is an optimization method for parallel simulation of natural genetic mechanisms and biological evolution and the formation of a random search, survival of the fittest principle of biological evolution exist in nature, was introduced to optimize the parameters of a coded series groups.

Orthogonal variable spreading factor (OVSF) codes are widely used to provide variable data rates for supporting different bandwidth requirements in wideband code division multiple access (WCDMA) systems. Many works in the literature have intensively investigated to find an optimal dynamic code assignment scheme for OVSF codes. Unlike earlier studies, which assign OVSF codes using conventional (CCA) or dynamic (DCA) code allocation schemes, in this paper, adaptive simulated annealing genetic algorithm (ASAGA) was applied which population is adaptively constructed according to existing traffic density in the OVSF code-tree. Also, the influences of the ASAGA parameters (selection, crossover and mutation techniques and cooling schedules) were examined on the dynamic OVSF code allocation problem. The simulation results show that the ASAGA provides reduced code blocking probability and improved spectral efficiency in the system when compared to the CCA and DCA schemes. ASAGA is also tested with its components SA and GA.

In this paper the application of a Genetic Algorithm (GA) to the Steiner tree problem is described. The performance of the GA is compared to that of the Simulated Annealing Algorithm (SA) and one of the best conventional algorithms given by Rayward-Smith and Clare [1] (RCA).Particular attention has been paid to find an optimal setting of the parameters and operators of the GA. A mutation probability P M=0.01 and a crossover probability P C=0.5 have been obtained according to the values found by Grefenstette [2]. An optimal population size has been chosen using a heuristic comparable to that of GoldbergIn addition the application of two problem specific heuristics is discussed, the removal of Steiner points of degree <3 and a local relaxation of the tree. A speedup of ≈20 was obtained if these heuristics were applied to the computation of the fitness only without changing the individuals themselves. For a fair comparison, these heuristics have been applied to the GA, the SA, and, in contrast to its original definition, to the RCA.According to our results, all three algorithms find the optimum and converge equally fast, i.e., GA and SA need not more function evaluations than the RCA, that has even been improved by the relaxation heuristic. The GA reaches the optimum in a small number of generations which is considered the reason why it did not show an even better performance.KeywordsGenetic AlgorithmFunction EvaluationConvergence SpeedSteiner TreeSimulated Annealing AlgorithmThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

The Genetic Algorithm (GA) is an explore technique used to solve issues in many different applications. The genetic algorithm has some parameters, including crossover probability, selection mechanism, and mutation probability. In GA, parameter adaptation is an important research topic. This paper proposes a Probabilistic Adaptive Genetic Algorithm in which the mutation and crossover probabilities, as well as the selection mechanism are dynamically adapted throughout the running of the algorithm. A new set of rates is generated for the next iteration based on the differences between fitness values and individual, enhancing the searching global optimum exploitation. We have compared the proposed algorithm with some common and state-of-the-art adaptive strategies such as dynamic adaptive, dynamic deterministic, dynamic self-adaptive, and static on a set of several functions with varying degrees of complexity. Experimental results on several popular test functions have shown that the results of the proposed algorithm are significantly better than these methods on both convergence speed and the solutions' quality.The reason that the proposed method has better results than other methods is the adaptation of each parameter of the genetic algorithm.

The query cost usually as an important criterion for a distributed database. The genetic algorithm is an adaptive probabilistic search algorithm, but the crossover and mutation probability usually keep a probability in traditional genetic algorithm. If the crossover probability keep a large value, the possibility of damage for genetic algorithm model is greater; In turn, if the crossover probability keep a small value, the search process will transform a slow processing or even stagnating. If the mutation probability keep a small value, a new individual can be reproduced difficultly; In turn, if the mutation probability keep a large value, the genetic algorithm will as a Pure random search algorithm. To solve this problem, proposed a improved genetic algorithm that multiple possibility of crossover and mutation based on k-means clustering algorithm. The experiment results indicate that the algorithm is effective.

Aiming at the time-varying and nonlinear problems of fixed-wing unmanned aerial vehicle (UAV) system, a UAV flight control algorithm was proposed based on PID control algorithm and the adaptive genetic algorithm (GA). On the basis of the traditional algorithm, the settling time was considered as the influencing factors of the objective function. The optimal index was determined by using the overshoot, rise time and settling time of the system response as variables, and adjusting the corresponding weight according to the actual situation to change the system response result. Moreover, the crossover probability and mutation probability in GA could be adaptively changed, so that the optimal parameters in the PID control process could be quickly and accurately found. Based on the rigid body dynamics model and kinematic model of the fixed-wing UAV, the simulation models of five control loops were designed in MATLAB/Simulink, namely the pitch angle, roll angle, yaw angle, speed and height control loops. Compared with the traditional PID controller, the system responses of the five control loops were significantly improved with faster response speed, and less than 5% overshoot. The proposed algorithm had strong adaptability and anti-interference ability.

Uncertainties in the production process would inevitably result in deviations from the existing schedule in flexible manufacturing systems (FMSs). This paper tries to study the problem in an environment with realistic interruptions and a requirement of time-restricted response in rescheduling. In our paper, the rescheduling system is based on the records of a dynamic database (DDB). It is able to reform the up-to-date status of a disturbed system via summarizing the remaining resources and works in process (WIPs) precisely. By using it as the new initial state, the new schedule is configured smoothly in conjunction with the existing schedule to improve the efficiency of FMS at this critical instant. Considering both speed and economic benefit, an adaptive genetic algorithm (AGA) is proposed for finding the new sub-optimal schedule of a large and complicated job shop FMS shortly after the interruption occurred. The AGA is designed to prevent the premature convergence and refine the performance of genetic algorithms in re-scheduling. During the AGA evolution process, the probabilities of crossover and mutation are varied, depending on both the fitness value and the normalized fitness distances between solutions. With help from DDB, the FMS scheduling model and AGA, the results obtained in the test examples of rescheduling are quite satisfactory.

What genetic algorithm's application process determined is a crossover probability and mutation probability. Based on the basis of the predecessor's study, we proposed adaptive crossover operator and mutation, and proved the convergence of the genetic algorithm is thus generated. The proposed adaptive genetic algorithms can be used to get fuzzy rules and criminal investigations in data mining, next to this method and research on digital forensics and criminal investigations.

In this paper, gratings to apply for the optical interconnection are designed using a genetic algorithm (GA) for a robust and efficient schema. The real-time optical interconnection system architecture is composed with LC-SLM, CCD array detector, IBM-PC, He-Ne laser, and Fourier transform lens. A pixelated binary phase grating is displayed on LC-SLM and could interconnect incoming beams to desired output spots freely by real-time. So as to adapt a GA for finding near globally-cost solutions, a chromosome is coded as a binary integer of length 32 X 32, the stochastic tournament method for decreasing the stochastic sampling error is performed, and a single-point crossover having 16 X 16 block size is used. The characteristics on the several parameters are analyzed in the desired grating design. Firstly, as the analysis of the effect on the probability of crossover, a designed grating when the probability of crossover is 0.75 has a 74.7[%] high diffraction efficiency and a 1.73 X 10<SUP>-1</SUP> uniformity quantitatively, where the probability of mutation is 0.001 and the population size is 300. Secondly, on the probability of mutation, a designed grating when the probability of mutation is 0.001 has a 74.4[%] high efficiency and a 1.61 X 10<SUP>-1</SUP> uniformity quantitatively, where the probability of crossover is 1.0 and the population size is 300. Thirdly, on the population size, a designed grating when the population size is 300 and the generation is 400 has above 74[%] diffraction efficiency, where the probability of mutation is 0.001 and the probability of crossover is 1.0.

Optimization model is build for solving the aircraft departure sequencing problem in this paper first. Then, an improved genetic algorithm (GA) using symbolic coding is proposed, where a type of total probability crossover and big probability mutation are performed. In this way, the evolutionary policy of Particle Swarm Optimization (PSO) is absorbed into the improved GA, which reduces the complexity and enhance the efficiency greatly. Last, a simulation program using basic GA, adaptive GA, and improved GA is performed. The simulation result shows that the model is effective and Improved GA has better performance than Basic GA or Adaptive GA.

The key parameters for Simulated Annealing Algorithm (SAA) and Genetic Algorithm (GA) were determined by test experiments. With optimal parameters obtained for both algorithms, the multilayer antireflection coatings in the 7.7-12.3μm band and a neutral beam splitter in the visible region were designed. Results showed that, for infrared multilayer antireflection coatings, the evaluation function value of SAA was 18.7% lower than that of GA, and the average transmittance of SAA was 0.4% higher than that of GA. Compared to GA, SAA achieved a higher average transmittance and a more flat transmittance curve. And for the neutral beam splitter, the running time of GA was 0.407s less than that of SAA, whereas the convergence effect of SAA was superior to that of GA. Therefore, the design with SAA has certain advantages for both the infrared multilayer antireflection coating and the beam splitter.

A combined strategy of an improved simulated annealing (SA) algorithm and genetic algorithm is presented, with the goal of reducing the computational expenses. The improvements made on the SA algorithm include two parts, i.e., the adaptive regulating for the step vector, and the dynamic testing for the equilibrium of the Metropolis process. The proposed strategy has both the advantage of SA algorithm, the ability to avoid being trapped in a local optimum, and that of genetic algorithm, the ability to use the information about the searched states for the next iteration. A practical application on geometry optimization of pole shoes in large salient pole synchronous generators is effectively implemented using the strategy. The numerical results show that the number of iterations used by executing the combined strategy are only about 75% of those by executing basic SA algorithm.

In Software Product Line (SPL) testing, reduced test suite with high coverage is useful for early features interaction detection. sGA (simplified genetic algorithm) and SAGA(simulated annealing genetic algorithm) can generate high coverage test suite. However, small probability mutations in updating test suite may reduce search efficiency and thus miss better solutions. An improved test cases generation method based on ASAGA (Adaptive simulated annealing genetic algorithm) is proposed. Experiments on SPLOT (Software Product Lines Online Tools) feature models show that the proposed hybrid ASAGA method can ensure local optimization accuracy and achieve smaller-size test suite with higher coverage.