Xiaofang Chen

A Hybrid Multiobjective Differential Evolution Algorithm and Its Application to the Optimization of Grinding and Classification

The grinding-classification is the prerequisite process for full recovery of the nonrenewable minerals with both production quality and quantity objectives concerned. Its natural formulation is a constrained multiobjective optimization problem of complex expression since the process is composed of one grinding machine and two classification machines. In this paper, a hybrid differential evolution (DE) algorithm with multi-population is proposed. Some infeasible solutions with better performance are allowed to be saved, and they participate randomly in the evolution. In order to exploit the meaningful infeasible solutions, a functionally partitioned multi-population mechanism is designed to find an optimal solution from all possible directions. Meanwhile, a simplex method for local search is inserted into the evolution process to enhance the searching strategy in the optimization process. Simulation results from the test of some benchmark problems indicate that the proposed algorithm tends to converge quickly and effectively to the Pareto frontier with better distribution. Finally, the proposed algorithm is applied to solve a multiobjective optimization model of a grinding and classification process. Based on the technique for order performance by similarity to ideal solution (TOPSIS), the satisfactory solution is obtained by using a decision-making method for multiple attributes.

Multi-step optimal control of complex process: a genetic programming strategy and its application

Abstract In many industrial processes, especially chemistry and metallurgy industry, the plant is slow for feedback and data test because of complex and varying factors. Considering the multi-objective feature and the complex problem of production stability in optimal control, this paper proposed an optimal control strategy based on genetic programming (GP), used as a multi-step state transferring procedure. The fitness function is computed by multi-step comprehensive evaluation algorithm, which provides a synthetic evaluation of multi-objective in process state based on single objective models. The punishment to process state variance is also introduced for the balance between optimal performance and stability of production. The individuals in GP are constructed as a chain linked by a few relation operators of time sequence for a facilitated evolution in GP with compact individuals. The optimal solution gained by evolution is a multi-step command program of process control, which not only ensures the optimization tendency but also avoids violent process variation by adjusting control parameters step by step. An optimal control system for operation direction is developed based on this strategy for imperial smelting process in Shaoguan. The simulation and application results showed its effectiveness for production objects optimization in complex process control.

Heterogeneous cooperative belief for social dilemma in multi-agent system

In the multi-agent system, there exist many agents, which are often heterogeneous, so the same decision-making protocol is not suitable for all the agents. In addition, it is generally acknowledged that when agents update their behavior, they often have a prior cooperative belief. Motivated by these observations, we put forward a novel behavior learning strategy, which takes the cooperative belief distribution and imitation dynamics into account to solve the social dilemma in multi-agent system. By conducting large-scale Monte Carlo simulations, we can easily draw a conclusion that the proposed behavior learning strategy can promote cooperation efficiently. In detail, a larger weight of cooperative belief is more beneficial to solving the social dilemma when η is in a suitable range. Especially, when the weight of cooperative belief is large enough, the cooperative agents can overcome the negative feedback mechanism introduced by network reciprocity, and make cooperation be the dominating behavior directly. In addition, when the value of η is larger than the threshold, the cooperation promotion effect is not straightforward. Therefore, when confronted with agents with heterogeneous cooperative belief, we should balance the cooperative belief and imitation dynamics in the behavior learning strategy to purse the optimal cooperation phase.

Modeling and Optimization Problems and Challenges Arising in Nonferrous Metallurgical Processes

Abstract Challenges in current development of nonferrous metallurgical industry include resource shortage, energy crisis and environmental pollution. The modeling and optimization are key techniques extensively used to save energy, reduce consumption and emissions in the nonferrous metallurgical processes. In this paper, firstly, the modeling problem for nonferrous metallurgical processes is considered. Based on the characteristics of the nonferrous metallurgical processes, several methods and theories for the modeling of nonferrous metallurgical processes, including the mechanism-based, continuous stirred tank reactor (CSTR)-based, and intelligent integrated modeling methods, are investigated. We focus on the description method in intelligent integrated modeling and its integration structures, and give some types of intelligent integrated models in various industrial applications. Secondly, the engineering optimization problem arising in nonferrous metallurgical processes is considered. Some engineering optimization methods, including operational-pattern optimization, satisfactory optimization with soft constraints adjustment, multi-objective intelligent optimization methods, and a comprehensive optimal control technique for a large-scale zinc electrolysis process are illustrated. In the end, some new challenges in process modeling and optimization are discussed.

Hybrid Intelligence Model Based on Image Features for the Prediction of Flotation Concentrate Grade

In flotation processes, concentrate grade is the key production index but is difficult to be measured online. The mechanism models reflect the basic tendency of concentrate grade changes but cannot provide adequate prediction precision. The data-driven models based on froth image features provide accurate prediction within well-sampled space but rely heavily on sample data with less generalization capability. So, a hybrid intelligent model combining the two kinds of model is proposed in this paper. Since the information of image features is enormous, and the relationship between image features and concentrate grade is nonlinear, a B-spline partial least squares (BS-PLS) method is adopted to construct the data-driven model for concentrate grade prediction. In order to gain better generalization capability and prediction accuracy, information entropy is introduced to integrate the mechanism model and the BS-PLS model together and modify the model output online through an output deviation compensation strategy. Moreover, a slide window scheme is employed to update the hybrid model in order to improve its adaptability. The industrial practical data testing results show that the performance of the hybrid model is better than either of the two single models and it satisfies the accuracy and stability requirements in industrial applications.

A PREDICTION MODEL OF SULFUR * BASED ON INTELLIGENT INTEGRATED STRATEGY

Abstract Considering the prediction of sulfur in agglomerate in sintering process, a prediction model based on intelligent integrated strategy is put forward in this paper, where the mathematical model calculates the sulfur content in agglomerate following material balance equation with some parameters predicted by NN method, while the expert rule model describes the relationship between sulfur quantity and key factors. An intelligent coordinator based on fuzzy logic is proposed to synthesize the output of the models. The industrial application proved its effectiveness in sintering production.

Anode effect prediction based on collaborative two-dimensional forecast model in aluminum electrolysis production

In this study, a new prediction algorithm is proposed, based on the collaborative two-dimensional forecast model (CTFM) that combines the traditional method and similarity search technique. The main idea of the algorithm is that the prediction of the change trend of the slope and the accumulated slope of the cell resistance as well as the useful knowledge obtained using the similarity search technique are used as the main criteria to calculate anode effect (AE)-prediction reliability. The accumulated mass deviation value is used as an auxiliary criterion to adjust the AE-prediction reliability. Finally, the current AE-process is marked according to the current AE-prediction reliability. The prediction model based on CTFM is tested on a real situation, in which multiple samples are extracted from the production of a 400 kA aluminum electrolysis cell. We observe that when the time advance of AE-prediction is within 20 ~ 40 min, the accuracy rate of the CTFM algorithm is greater than 95% and the applicability of the method is excellent, showing a high prediction accuracy for different aluminum electrolysis cells.

A knowledge reasoning Fuzzy-Bayesian network for root cause analysis of abnormal aluminum electrolysis cell condition

Root cause analysis (RCA) of abnormal aluminum electrolysis cell condition has long been a challenging industrial issue due to its inherent complexity in analyzing based on multi-source knowledge. In addition, accurate RCA of abnormal aluminum electrolysis cell condition is the precondition of improving current efficiency. RCA of abnormal condition is a complex work of multi-source knowledge fusion, which is difficult to ensure the RCA accuracy of abnormal cell condition because of dwindling and frequent flow of experienced technicians. In view of this, a method based on Fuzzy-Bayesian network to construct multi-source knowledge solidification reasoning model is proposed. The method can effectively fuse and solidify the knowledge, which is used to analyze the cause of abnormal condition by technicians providing a clear and intuitive framework to this complex task, and also achieve the result of root cause automatically. The proposed method was verified under 20 sets of abnormal cell conditions, and implements root cause analysis by finding the abnormal state of root node, which has a maximum posterior probability by Bayesian diagnosis reasoning. The accuracy of the test results is up to 95%, which shows that the knowledge reasoning feasibility for RCA of aluminum electrolysis cell.

Adaptive Image Processing for Bubbles in Flotation Process

In this paper, an adaptive segmentation based image processing method is presented for mineral flotation processes so as to obtain an optimal quality profile measure. An optimal structural element (SE) is firstly provided for initial image segmentation where fuzzy c-means (FCM) algorithm is combined with a watershed algorithm. Then, the extracted fuzzy texture spectrum feature is discriminated as either under-segmentation or over-segmentation regions. Finally, the segmentation results are subsequently processed to estimate the output probability density function (PDF) of bubble size distribution, which is the basis of the fault detection and real-time control for flotation processes. Experimental results demonstrate that the proposed method avoids both over-segmentation and under segmentation, and reveal the reliability of bubble size distribution nonparametric density estimation.