The single channel Convolutional Neural Network (CNN) cannot fully study the feature information of text with a single perspective. In order to solve the problem, a new Dual-Channel CNN (DCCNN) algorithm was proposed. Firstly, the word vector was trained by Word2Vec, and the semantic information of sentence was obtained by using word vector. Secondly, two different channels were used to carry out convolution operations, one channel was the character vector and the other was the word vector. The fine-grained character vector was used for assisting word vector to capture deep semantic information. Finally, the convolutional kernels of different sizes were used to find higher-level abstract features within the sentence. The experimental results show that, the proposed DCCNN algorithm can accurately identify the sentiment polarity of text, its accuracy and F1 value are above 95%, which are significantly improved compared with the algorithms of logistic regression, Support Vector Machine (SVM) and CNN.

To overcome the demerits of basic Shuffled Frog Leaping Algorithm (SFLA), such as slow convergence speed, low optimization precision and falling into local optimum easily, a novel survival of the fittest SFLA with normal mutation was proposed. In the local search strategy of the proposed algorithm, the normal mutations for updating strategy of the worst frog individuals in the subgroup were introduced to avoid the algorithm falling into local convergence effectively, expand the searching space and increase the diversity of population. Meanwhile, the mutations were selected for a small number of worse frog individual in the subgroup to inherit the useful mutations instead of the bad mutations. The survival of the fittest was implemented, the quality of the population was improved, the blindness of the algorithm optimization process was reduced and the algorithm optimization was speeded up. The elite mutation mechanism for the best frog individuals in each subgroup was introduced for obtaining better individuals to enhance the global optimization ability of the algorithm further, avoid falling into local convergence, and lead the whole population evolution to the better. The experimental results of 30 independent runs indicate that the proposed algorithm can converge to the optimal solution of 0 in Sphere, Rastrigrin, Griewank, Ackley and Quadric, which is better than the other contrastive algorithms. The experimental results show that the proposed algorithm can avoid falling into premature convergence effectively, improve the convergence speed and convergence precision.

To solve multi-objective Flexible Job-shop Scheduling Problems (FJSP), a Teaching and Peer-Learning Particle Swarm Optimization with Pareto Non-Dominated Solution Set (PNDSS-TPLPSO) algorithm was proposed. First, the minimum completion time of jobs, the maximum work load of machines and the total work load of all machines were taken as the optimization goals to establish a multi-objective flexible job-shop scheduling model. Then, the proposed algorithm combined multi-objective Pareto method with Teaching and Peer-Learning Particle Swarm Optimization (TPLPSO). A fast Pareto non-dominated sorting operator was applied to generate initial Pareto non-dominated solution set, and extracting Pareto dominance layer program was adopted to update Pareto non-dominated solution set. Furthermore, composite dispatching rule was adopted to generate the initial population, and opening up parabola decreasing inertia weigh strategy was taken to improve the convergence speed. Finally, the proposed algorithm was adopted to solve three Benchmark instances. In the comparison experiments with Multi-Objective Evolutionary Algorithm with Guided Local Search (MOEA-GLS) and Controlled Genetic Algorithm with Approach by Localization (AL-CGA), the proposed algorithm can obtain more and better Pareto non-dominated solutions for the same Benchmark instance. In terms of computing time, the proposed algorithm is less than MOEA-GLS. The simulation results demonstrate that the proposed algorithm can solve multi-objective FJSP effectively.

To deal with the Flexible Job-shop Scheduling Problem (FJSP), an Improved Two-Layer Particle Swarm Optimization (ITLPSO) algorithm was proposed. First, minimization of the maximal completion time of all machines was taken as the optimization objective to establish a flexible job-shop scheduling model. And then the improved two-layer PSO algorithm was presented, in which the stagnation prevention strategy and concave function decreasing strategy were adopted to avoid falling into local optimum and to improve the convergence rate. Finally, the proposed algorithm was adopted to solve the relevant instance and the comparison with existing methods was also performed. The experimental results showed that, compared with the standard PSO algorithm and the Two-Layer Particle Swarm Optimization (TLPSO) algorithm, the optimal value of the maximum completion time was reduced by 11 and 6 respectively, the average maximum completion time was reduced by 15.7 and 4 respectively, and the convergence rate was improved obviously. The performance analysis shows that the proposed algorithm can improve the efficiency of the flexible job-shop scheduling obviously and obtain better scheduling solution.

Concerning the problem that Particle Swarm Optimization (PSO) falls into local minima easily and converges slowly at the last stage, a kind of hybrid PSO algorithm with cooperation of multiple particle roles (MPRPSO) was proposed. The concept of particle roles was introduced into the algorithm to divide the population into three roles: Exploring Particle (EP), Patrolling Particle (PP) and Local Exploiting Particle (LEP). In each iteration, EP was used to search the solution space by the standard PSO algorithm, and then PP which was based on chaos was used to strengthen the global search capability and replace some EPs to restore population vitality when the algorithm trapped in local optimum. Finally, LEP was used to strengthen the local search to accelerate convergence by unidimensional asynchronous neighborhood search. The 30 times independent runs in the experiment show that, the proposed algorithm in the conditions that particle roles ratio is 0.8∶〖KG-*3〗0.1∶〖KG-*3〗0.1 has the mean value of 2.352E-72,4.678E-29,7.780E-14 and 2.909E-14 respectively in Sphere, Rosenbrock, Ackley and Quadric, and can converge to the optimal solution of 0 in Rastrigrin and Griewank, which is better than the other contrastive algorithms. The experimental results show that proposed algorithm improves the optimal performance with certain robustness.