Unit commitment is one of the most important subjects with respect to the economical operation of power systems, which attempts to minimize the total thermal generating cost while satisfying all the necessary restrictive conditions. ¡@¡@This thesis proposes a short-term thermal generating unit commitment by genetic algorithm and back propagation network. Genetic algorithm is based on the optimization theory developed from natural evolution principles, and in the optimization process, seeks a set of solutions simultaneously rather than any single one by adopting stochastic movement rule from one solution to another, which prevents restriction to fractional minimal values. Neural networks method outperforms in speed and stability. This thesis uses back propagation network method to complete neural networks and sets the optimal unit combination derived from genetic algorithm as the target output. ¡@¡@Under fixed electrical systems, instant responsiveness can be calculated by neural networks. When the systematical architecture changes, genetic algorithm can be applied to re-evaluation of the optimal unit commitment, hoping to improve the pitfalls of traditional methods. ¡@¡@This thesis takes the power system of six units for example to conduct performance assessment. The results show that genetic algorithm provides solutions closer to the overall optimal solution than traditional methods in optimizing unit commitment. On the other hand, neural networks method can not only approximate the solution obtained by genetic algorithm but also process faster than any other methods.

Building a rule-based classification system for a training data set is an important research topic in the area of data mining, knowledge discovery and expert systems. Recently, the GA-based fuzzy approach is shown to be an effective way to design an efficient evolutionary fuzzy system. In this thesis a three layers genetic algorithm with Simulated Annealing for selecting a small number of fuzzy if-then rules to building a compact fuzzy classification system will be proposed. The rule selection problem with three objectives: (1) maximize the number of correctly classified patterns, (2) minimize the number of fuzzy if-then rules, and (3) minimize the number of required features. Genetic algorithms are applied to solve this problem. A set of fuzzy if-then rules is coded into a binary string and treated as an in-dividual in genetic algorithms. The fitness of each individual is specified by three ob-jectives in the combinatorial optimization problem. Simulated annealing (SA) is op-tionally cooperated with three layers genetic algorithm to effectively select some layer control genes. The performance of the proposed method for training data and test data is ex-amined by computer simulations on the iris data set and spiral data set, and comparing the performance with the existing approaches. It is shown empirically that the pro-posed method outperforms the existing methods in the design of optimal fuzzy sys-tems.

When we use the FDTD method to analyze electromagnetic problems, it has to properly discretize the space and time. Automeshing can non-uniformly discretize the simulated structure and generate gradual grids. To improve the efficiency of automeshing, we optimize the parameter of automeshing using the genetic algorithm. Without sacrificing accuracy, it searches a suitable ratio to reduce the generated grids and to save simulation time. At last, we optimize the PIFA using genetic algorithm and search automatically the height of the substrate and the feed position in order to obtain optimal performance. When we use the genetic algorithm, it is the key point to define an objective function evaluating the fitness of the optimized problem. It is important that the function has to appropriately describe the performance at that time.

In airline industry scheduling problems, the aircraft routing and the aircrew pairing problems are highly related to fueling and personnel costs. When performing aircraft routing and aircrew pairing, several objectives, such as the ground-turn around time, flow balance, transition time, number of deadheads, number of layovers, flying time, and flight duty period should be considered. It is difficult to optimize these conflicting objectives simultaneously. Many issues are yet to be solved as follows. 1. Most researches related to the aircraft routing and aircrew pairing problems use set partitioning or set covering models. Planners must (1) enumerate several possible subsets of flights, (2) assign costs, and (3) check feasibilities simultaneously. This is time-consuming since the numbers of whole subsets are exponential values to the problem size. 2. The number of enumerated subsets is usually too small to cover the whole solution space. Therefore, even if the optimal solution is found, it is just a local optimal solution of the enumerated subsets. 3. When using traditional optimization algorithms to find a combination of these subsets with minimal cost, it should be ensured that all flights should be covered exactly once. This causes the overheads of checking the number of coverage. 4. In traditional solution methods, the number of required aircrafts and crewmembers cannot be pre-specified since these numbers can only be obtained when the optimization algorithm is completed. 5. All enumerated subsets should be assigned cost values according to various objectives, such as transition time, number of deadheads, number of layovers, flying time, and flight duty period. The cost values are difficult to assign since it is highly dependent on domain knowledge, and usually nonlinear. Also, inappropriate cost values will cause bias in optimization, and ambiguity among all factors due to single objective formulation. Hence, to overcome these problems, we propose several enhancements in both formulation and the solution stages. In the formulation stage, we propose a novel permutation-based model with multiple objectives, which has the following features. 1. The proposed permutation-based model can save the overheads of pre-enumerating possible sub-solutions 2. The permutation-based model can cover the whole solution space. Hence, it has more chance to find out the global optimal solution. 3. The proposed permutation-based model can ensure that each flight can be covered exactly once to save the overheads of checking the number of coverage. 4. The proposed permutation-based model can provide a new way to pre-specify the number of aircrafts or group number of crewmembers. 5. Taking the advantage of multiobjective formulation, various objectives are considered separatively instead of assigning cost values. All objective can be considered individually even if they have different definitions of optimality or scales. In the solution stage, we apply the MOI-based MGA (MMGA) to solve the problems of aircraft routing and crew pairing. MMGA is originally proposed to solve numerical controller design problem. By using MMGA, designers can configure the ranges of solutions via adjusting an auxiliary vector of performance indices. To make MMGA more suitable for solving the aircraft routing and aircrew pairing problems, some enhancements are added, such as chromosome encoding scheme, repairing strategy, crossover, and mutation operations. This approach has following features. 1. In both aircraft routing and aircrew pairing problems, the permutation-based encoding scheme, which is the same as the formulation model, can ensure all flights be covered once. 2. Moreover, in the crew pairing problem, the sectional permutation-based encoding scheme, which divides the flights into three sections, such as earlier flights, later flights, and floating flights, can enhance MMGA to find out optimal solutions which satisfy the flight duty period objective. 3. Also, to overcome the large violations caused by random generation of candidate solutions, we use a repairing strategy, which repairs all generated solutions by reordering the sequences of flights according to departure times. 4. The sectional order-based crossover can have a more stable evolution than the widely-used partial mapped crossover. Also, it can make the newborn offspring keep the features of three sections defined in the encoding scheme. 5. Also, the sectional mutation can inherit the advantages of the widely-used reciprocal mutation and keep the features of three sections defined in the encoding scheme. In the aircraft routing problem, experiments show that MMGA can find out optimal flight schedules under the condition of sufficient aircrafts. On the other aspect, when the number of aircrafts is insufficient, planners can modify the obtained solutions by a little retiming process when the number of violations is small. In the aircrew pairing problem, experiments indicate the proposed approach can solve the aircrew pairing problem with minimal group number of crewmembers which is verified by a branch-and-bound approach. By using MMGA, the problems of aircraft routing and aircrew pairing can be solved efficiently and effectively. In other words, planners can solve these problems in a short time period instead of enumeration and feasibility checking by traditional methods. Via the proposed approach, planners can further consider more important issues, such as to suggest better schedules with lower cost and higher benefit.

The microarray technology plays an important role of clinical oncology field. The patient can be diagnosed a symptom about cancer through microarray data. Currently, to solve classification of microarray data is still a wild open issue. Existing methods may have a good performance, but need to spend much time to analyze microarray data, such as SVM. In this thesis, we propose a novel GA-Fuzzy-based voting mechanism to find genes which affect the symptom to better diagnose patient. The proposed algorithm can blur the boundary between classes to handle the ambiguous regions. In order to simulate the gene selection mechanism, we proposed upper bound £\-Cut and lower bound £\-Cut in voting mechanism. Two groups of data collected from the literature are used to test the performance of the proposed algorithm. In the first group of dataset, experimental results show that the accuracies of five datasets using the proposed algorithm are better than those methods proposed by Pochet et al. But, there are the four datasets which the accuracies using the proposed algorithm are a little bit worse than the methods proposed by Pochet et al. For the second group of dataset, the accuracies of seven datasets using the proposed algorithm are better than KerNN proposed by Xiong and Chen. But, there are four datasets which the accuracies using the proposed algorithm are worse than KerNN proposed by Xiong and Chen. Nevertheless, experimental results show that the proposed algorithm performs the best for multi-class data.

Various approaches have been proposed for finding the longest common subsequence (LCS) of two sequences. The time complexities of these algorithms are usually $O(n^2)$ in the worst case, where $n$ is the length of input sequences. However, these algorithms would become infeasible when the input length, $n$, is very long. Recently, the $k$-LCS $(k ≥ 2)$ problem has become more attractive. Some algorithms have been proposed for solving the problem, but the execution time required for solving the $k$-LCS problem is still too long to be practical. In this thesis, we propose a genetic algorithm for solving the $k$-LCS problem with time complexity $O(Gpk(n + |P_j|))$, which $G$ is the number of generations, $p$ is the number of template patterns, $k$ is the number of input sequences, $n$ and $|P_j|$ are the length of input sequences and the length of template patterns, respectively. As our experimental results show, when $k$ is 20 and $n$ is 1000, the performance ratio ($|CS|/|LCS|$) of our algorithm is greater than 0.8, where $|CS|$ denotes the length of the solution we find, and $|LCS|$ represents the length of the real (optimal) LCS. Comparing the performance ratios with Expansion Algorithm and BNMAS Algorithm, our algorithm is much better than them when the number of input sequences varies from 2 to 20 and the length of the input sequences varies from 100 to 2000.

In this thesis, the Hybrid Genetic Algorithm-Ant Colony Optimization (GACO) approach is presented to solve the unit commitment problem (UC), and comparison with the results obtained using literature methods. Then this thesis applied the ability of the Genetic Algorithm (GA) operated after Ant Colony Optimization (ACO) can promote the ACO efficiency. The objective of GA is to improve the searching quality of ants by optimizing themselves to generate a better result, because the ants produced randomly by pheromone process are not necessary better. This method can not only enhance the neighborhood search, but can also search the optimum solution quickly to advance convergence. The other objective of this thesis is to investigate an influence of emission constraints on generation scheduling. The motivation for this objective comes from the efforts to reduce negative trends in a climate change. In this market structure, the independent power producers have to deal with several complex issues arising from uncertainties in spot market prices, and technical constraints which need to be considered while scheduling generation and trading for the next day. In addition to finding dispatch and unit commitment decisions while maximizing its profit, their scheduling models should include trading decisions like spot-market buy and sell. The model proposed in this thesis build on the combined carbon finance and spot market formulation, and help generators in deciding on when these commitments could be beneficial.

This thesis presents an imitation-based methodology, also a simple and easy way, for a service robot to learn the behaviors demonstrated by the user. With this proposed method, a robot can learn human behavior through observation. Inspired by the concept of biological learning, this learning model is initiated when facing a new learning event. A series of experiments are conducted to use a humanoid robot as a platform to implement the proposed algorithm. Discussions are made of how the robot generates a complete behavior sequences performed by its demonstrator. Because it is time consuming for a robot to go through the whole process of learning, we thus propose a decomposed learning method to enhance the learning performance, that is, based on the past learning information, the robot can skip learning again the behaviors already known. For simple robot behaviors, a hierarchical evolutionary mechanism is developed to evolve the complete behavior trajectories. For complex behaviors sequences, different ways are used to tackle the scalability problem, including decomposing the overall task into several sub-tasks, exploiting behavior information recorded previously, and constructing a new strategy to maintain population diversity. To verify our approach, a different series of experiments have been conducted. The results show that our imitation-based approach is a natural way to teach the robot new behaviors. This evolutionary mechanism successfully enables a humanoid robot to perform the behavior sequences it learns.

The domain of computer music is an interesting area which combines computer science and music art. We propose a music accompaniment system using a variable-length genetic algorithm. Via the system one can make the music corresponded to his demands. In the style music accompaniment we analyze some important characteristic of pop music, and propose a new chromosome representation scheme to include the concept of rhyme, chord and melody. Chord progression is used as one of the evaluation criterions in this thesis. The system allows a user to input melody, to select emotion and rhyme, and the system will automatically generate the appropriate accompaniment based on the database compiled from some music theory relating to the chord progression. In addition, the system allows a user to select his favorite accompaniment that generated by the system. Based on the user selected accompaniment the system will generate similar accompaniments for the user.

The multiobjective genetic algorithm can be used to optimize two conflicting objectives, oil production and polymer utility factor in polymer flood design. This approach provides a set of optimal solutions which can be considered as trade-off curve (Pareto front) to maximize oil production while preserving polymer performance. Then an optimal polymer flood design can be considered from post-optimization analysis. A 2D synthetic example, and a 3D field-scale application, accounting for geologic uncertainty, showed that beyond the optimal design, a relatively minor increase in oil production requires much more polymer injection and the polymer utility factor increases substantially.