Optimization of linear induction machines based on a novel adaptive genetic algorithm

In order to improve the thrust-power ratio index of the linear induction motor（LIM）, a novel adaptive genetic algorithm （NAGA） is proposed for the design optimization of the LIM. A good-point set theory that helps to produce a uniform initial population is used to enhance the optimization efficiency of the genetic algorithm. The crossover and mutation probabilities are improved by using the function of sigmoid and they can be adjusted nonlinearly between average fitness and maximal fitness with individual fitness. Based on the analyses of different structures between the LIM and the rotary induction motor （RIM） and referring to the analysis method of the RIM, the steady-state characteristics of the LIM that considers the end effects of the LIM is calculated and the optimal design model of the thrust-power ratio index is also presented. Through the comparison between the optimal scheme and the old scheme, the thrust-power ratio index of the LIM is obviously increased and the validity of the NAGA is proved.

Classification of hyperspectral remote sensing images based on simulated annealing genetic algorithm and multiple instance learning

A hybrid feature selection and classification strategy was proposed based on the simulated annealing genetic algonthrn and multiple instance learning （MIL）. The band selection method was proposed from subspace decomposition, which combines the simulated annealing algorithm with the genetic algorithm in choosing different cross-over and mutation probabilities, as well as mutation individuals. Then MIL was combined with image segmentation, clustering and support vector machine algorithms to classify hyperspectral image. The experimental results show that this proposed method can get high classification accuracy of 93.13% at small training samples and the weaknesses of the conventional methods are overcome.

Identification of Crop Diseases Based on Improved Genetic Algorithm and Extreme Learning Machine

As an indispensable task in crop protection,the detection of crop diseases directly impacts the income of farmers.To address the problems of low crop-disease identification precision and detection abilities,a new method of detection is proposed based on improved genetic algorithm and extreme learning machine.Taking five different typical diseases with common crops as the objects,this method first preprocesses the images of crops and selects the optimal features for fusion.Then,it builds a model of crop disease identification for extreme learning machine,introduces the hill-climbing algorithm to improve the traditional genetic algorithm,optimizes the initial weights and thresholds of the machine,and acquires the approximately optimal solution.And finally,a data set of crop diseases is used for verification,demonstrating that,compared with several other common machine learning methods,this method can effectively improve the crop-disease identification precision and detection abilities and provide a basis for the identification of other crop diseases.

A Hybrid Genetic Algorithm for Supervised Inductive Learning

A novel algorithm is presented for supervised inductive learning by integrating a genetic algorithm with hot'tom-up induction process.The hybrid learning algorithm has been implemented in C on a personal computer(386DX/40).The performance of the algorithm has been evaluated by applying it to 11-multiplexer problem and the results show that the algorithm's accuracy is higher than the others[5,12, 13].

Learning Bayesian networks using genetic algorithm

A new method to evaluate the fitness of the Bayesian networks according to the observed data is provided. The main advantage of this criterion is that it is suitable for both the complete and incomplete cases while the others not. Moreover it facilitates the computation greatly. In order to reduce the search space, the notation of equivalent class proposed by David Chickering is adopted. Instead of using the method directly, the novel criterion, variable ordering, and equivalent class are combined,moreover the proposed mthod avoids some problems caused by the previous one. Later, the genetic algorithm which allows global convergence, lack in the most of the methods searching for Bayesian network is applied to search for a good model in thisspace. To speed up the convergence, the genetic algorithm is combined with the greedy algorithm. Finally, the simulation shows the validity of the proposed approach.

Energy Theft Detection in Smart Grids with Genetic Algorithm-Based Feature Selection

As big data,its technologies,and application continue to advance,the Smart Grid(SG)has become one of the most successful pervasive and fixed computing platforms that efficiently uses a data-driven approach and employs efficient information and communication technology(ICT)and cloud computing.As a result of the complicated architecture of cloud computing,the distinctive working of advanced metering infrastructures(AMI),and the use of sensitive data,it has become challenging tomake the SG secure.Faults of the SG are categorized into two main categories,Technical Losses(TLs)and Non-Technical Losses(NTLs).Hardware failure,communication issues,ohmic losses,and energy burnout during transmission and propagation of energy are TLs.NTL’s are human-induced errors for malicious purposes such as attacking sensitive data and electricity theft,along with tampering with AMI for bill reduction by fraudulent customers.This research proposes a data-driven methodology based on principles of computational intelligence as well as big data analysis to identify fraudulent customers based on their load profile.In our proposed methodology,a hybrid Genetic Algorithm and Support Vector Machine(GA-SVM)model has been used to extract the relevant subset of feature data from a large and unsupervised public smart grid project dataset in London,UK,for theft detection.A subset of 26 out of 71 features is obtained with a classification accuracy of 96.6%,compared to studies conducted on small and limited datasets.

Optimizing Deep Learning Parameters Using Genetic Algorithm for Object Recognition and Robot Grasping

The performance of deep learning(DL)networks has been increased by elaborating the network structures. However, the DL netowrks have many parameters, which have a lot of influence on the performance of the network. We propose a genetic algorithm(GA) based deep belief neural network(DBNN) method for robot object recognition and grasping purpose. This method optimizes the parameters of the DBNN method, such as the number of hidden units, the number of epochs, and the learning rates, which would reduce the error rate and the network training time of object recognition. After recognizing objects, the robot performs the pick-andplace operations. We build a database of six objects for experimental purpose. Experimental results demonstrate that our method outperforms on the optimized robot object recognition and grasping tasks.

Fuzzy-second order sliding mode control optimized by genetic algorithm applied in direct torque control of dual star induction motor

The direct torque control of the dual star induction motor(DTC-DSIM) using conventional PI controllers is characterized by unsatisfactory performance, such as high ripples of torque and flux, and sensitivity to parametric variations. Among the most evoked control strategies adopted in this field to overcome these drawbacks presented in classical drive, it is worth mentioning the use of the second order sliding mode control(SOSMC) based on the super twisting algorithm(STA) combined with the fuzzy logic control(FSOSMC). In order to realize the optimal control performance, the FSOSMC parameters are adjusted using an optimization algorithm based on the genetic algorithm(GA). The performances of the envisaged control scheme, called G-FSOSMC, are investigated against G-SOSMC, G-PI and BBO-FSOSMC algorithms. The proposed controller scheme is efficient in reducing the torque and flux ripples, and successfully suppresses chattering. The effects of parametric uncertainties do not affect system performance.

Optimization of total harmonic current distortion and torque pulsation reduction in high-power induction motors using genetic algorithms

This paper presents a powerful application of genetic algorithm （GA） for the minimization of the total harmonic current distortion （THCD） in high-power induction motors fed by voltage source inverters, based on an approximate harmonic model. That is, having defined a desired fundamental output voltage, optimal pulse patterns （switching angles） are determined to produce the fundamental output voltage while minimizing the THCD. The complete results for the two cases of three and five switching instants in the first quarter period of pulse width modulation （PWM） waveform are presented. Presence of harmonics in the stator excitation leads to a pulsing-torque component. Considering the fact that if the pulsing-torques are at low frequencies, they can cause troublesome speed fluctuations, shaft fatigue, and unsatisfactory performance in the feedback control system, the 5th, 7th, 1 lth, and 13th current harmonics （in the case of five switching angles） are constrained at some pre-specified values, to mitigate the detrimental effects of low-frequency harmonics. At the same time, the THCD is optimized while the required fundamental output voltage is maintained.

A Linkage Learning Genetic Algorithm with Linkage Matrix

The goal of linkage learning, or building block identification, is the creation of a more effective Genetic Algorithm （GA）. This paper proposes a new Linkage Learning Genetic Algorithms, named m-LLGA. With the linkage learning module and the linkage-based genetic operation, m-LLGA is not only able to learn and record the linkage information among genes without any prior knowledge of the function being optimized. It also can use the linkage information stored in the linkage matrix to guide the selection of crossover point. The preliminary experiments on two kinds of bounded difficulty problems and a TSP problem validated the performance of m-LLGA. The m-LLGA learns the linkage of different building blocks parallel and therefore solves these problems effectively; it can also reasonably reduce the probability of building blocks being disrupted by crossover at the same time give attention to getting away from local minimum.

Linkage intensity learning approach with genetic algorithm for causality diagram

The causality diagram theory, which adopts graphical expression of knowledge and direct intensity of causality, overcomes some shortages in belief network and has evolved into a mixed causality diagram methodology for discrete and continuous variable. But to give linkage intensity of causality diagram is difficult, particularly in many working conditions in which sampling data are limited or noisy. The classic learning algorithm is hard to be adopted. We used genetic algorithm to learn linkage intensity from limited data. The simulation results demonstrate that this algorithm is more suitable than the classic algorithm in the condition of sample shortage such as space shuttle’s fault diagnoisis.

SELF-LEARNING FUZZY CONTROL RULES USING GENETIC ALGORITHMS

This papcr presents a new genetic algorithms(GAs)-based method for self-learniag fuzzy control rules. An improved GA is used to learn to optimally select the fuzzy membership functions of the linguistic labels in the condition portion of each rule, and to automatically generate fuzzy control actions under each condition. The dynamics of the controlled system is unknown to the GA. The only information for evaluating performance is a failure signal indicating that the controlled system is out of control. We compare its performance with that of other learning methods for the same problem. We also examine the ability of the algorithm to adapt to changing conditions. Simulation results show that such an approach for self-learning fuzzy control rules is both effective and robust.

Self-learning Fuzzy Controllers Based On a Real-time Reinforcement Genetic Algorithm

This paper presents a novel method for constructing fuzzy controllers based on a real time reinforcement genetic algorithm. This methodology introduces the real-time learning capability of neural networks into globally searching process of genetic algorithm, aiming to enhance the convergence rate and real-time learning ability of genetic algorithm, which is then used to construct fuzzy controllers for complex dynamic systems without any knowledge about system dynamics and prior control experience. The cart-pole system is employed as a test bed to demonstrate the effectiveness of the proposed control scheme, and the robustness of the acquired fuzzy controller with comparable result.

Self-Organizing Genetic Algorithm Based Method for Constructing Bayesian Networks from Databases

The typical characteristic of the topology of Bayesian networks (BNs) is the interdependence among different nodes (variables), which makes it impossible to optimize one variable independently of others, and the learning of BNs structures by general genetic algorithms is liable to converge to local extremum. To resolve efficiently this problem, a self-organizing genetic algorithm (SGA) based method for constructing BNs from databases is presented. This method makes use of a self-organizing mechanism to develop a genetic algorithm that extended the crossover operator from one to two, providing mutual competition between them, even adjusting the numbers of parents in recombination (crossover/recomposition) schemes. With the K2 algorithm, this method also optimizes the genetic operators, and utilizes adequately the domain knowledge. As a result, with this method it is able to find a global optimum of the topology of BNs, avoiding premature convergence to local extremum. The experimental results proved to be and the convergence of the SGA was discussed.

Using a Hybrid Genetic Algorithm to Learn Rules From Examples

in this article a novel learning method is proposed,which is a combination of GA and the bottomup induction process. The method has been implemented in a system called KAA,and we evaluate it on a multiplexer problem,which shows the higher predict accuracy even in a noisy environment.

Prediction of Permeability Using Random Forest and Genetic Algorithm Model

Precise recovery of CoalbedMethane(CBM)based on transparent reconstruction of geological conditions is a branch of intelligent mining.The process of permeability reconstruction,ranging from data perception to real-time data visualization,is applicable to disaster risk warning and intelligent decision-making on gas drainage.In this study,a machine learning method integrating the Random Forest(RF)and the Genetic Algorithm(GA)was established for permeability prediction in the Xishan Coalfield based on Uniaxial Compressive Strength(UCS),effective stress,temperature and gas pressure.A total of 50 sets of data collected by a self-developed apparatus were used to generate datasets for training and validating models.Statistical measures including the coefficient of determination(R2)and Root Mean Square Error(RMSE)were selected to validate and compare the predictive performances of the single RF model and the hybrid RF–GA model.Furthermore,sensitivity studies were conducted to evaluate the importance of input parameters.The results show that,the proposed RF–GA model is robust in predicting the permeability;UCS is directly correlated to permeability,while all other inputs are inversely related to permeability;the effective stress exerts the greatest impact on permeability based on importance score,followed by the temperature(or gas pressure)and UCS.The partial dependence plots,indicative of marginal utility of each feature in permeability prediction,are in line with experimental results.Thus,the proposed hybrid model(RF–GA)is capable of predicting permeability and thus beneficial to precise CBMrecovery.

Rotation forest based on multimodal genetic algorithm

In machine learning,randomness is a crucial factor in the success of ensemble learning,and it can be injected into tree-based ensembles by rotating the feature space.However,it is a common practice to rotate the feature space randomly.Thus,a large number of trees are required to ensure the performance of the ensemble model.This random rotation method is theoretically feasible,but it requires massive computing resources,potentially restricting its applications.A multimodal genetic algorithm based rotation forest(MGARF)algorithm is proposed in this paper to solve this problem.It is a tree-based ensemble learning algorithm for classification,taking advantage of the characteristic of trees to inject randomness by feature rotation.However,this algorithm attempts to select a subset of more diverse and accurate base learners using the multimodal optimization method.The classification accuracy of the proposed MGARF algorithm was evaluated by comparing it with the original random forest and random rotation ensemble methods on 23 UCI classification datasets.Experimental results show that the MGARF method outperforms the other methods,and the number of base learners in MGARF models is much fewer.

Fuzzy Control of Chaotic System with Genetic Algorithm

A novel approach to control the unpredictable behavior of chaotic systems is presented. The control algorithm is based on fuzzy logic control technique combined with genetic algorithm. The use of fuzzy logic allows for the implementation of human "rule-of-thumb" approach to decision making by employing linguistic variables. An improved Genetic Algorithm (GA) is used to learn to optimally select the fuzzy membership functions of the linguistic labels in the condition portion of each rule, and to automatically generate fuzzy control actions under each condition. Simulation results show that such an approach for the control of chaotic systems is both effective and robust.

Amino Acid Encryption Method Using Genetic Algorithm for Key Generation

In this new information era,the transfer of data and information has become a very important matter.Transferred data must be kept secured from unauthorized persons using cryptography.The science of cryptography depends not only on complex mathematical models but also on encryption keys.Amino acid encryption is a promising model for data security.In this paper,we propose an amino acid encryption model with two encryption keys.The first key is generated randomly using the genetic algorithm.The second key is called the protein key which is generated from converting DNA to a protein message.Then,the protein message and the first key are used in the modified Playfair matrix to generate the cypher message.The experimental results show that the proposed model survives against known attacks such as the Brute-force attack and the Ciphertext-only attack.In addition,the proposed model has been tested over different types of characters including white spaces and special characters,as all the data is encoded to 8-bit binary.The performance of the proposed model is compared with other models using encryption time and decryption time.The model also balances all three principles in the CIA triad.

Predicting Optimal Trading Actions Using a Genetic Algorithm and Ensemble Method

Machine learning has been applied to the foreign exchange market for algorithmic trading. However, the selection of trading algorithms is a difficult problem. In this work, an approach that combines trading agents is designed. In the proposed approach, an artificial neural network is used to predict the optimum actions of each agent for USD/JPY currency pairs. The agents are trained using a genetic algorithm and are then combined using an ensemble method. We compare the performance of the combined agent to the average performance of many agents. Simulation results show that the total return is better when the combined agent is used.