Research On Computer Recognization Of Taste Signals

Machine vision, hearing, tactual sensation and force sensation are greatly developed in the domain of robotics and some of them have been used for practical purposes. Machine taste and smell sensation have wide applications in the intellectualized management in food industries, quality inspection of food, evaluation of taste and smell and so on. However, the progress made in machine taste and smell sensation are far from satisfaction. The most challenging tasks include to construct taste and smell sensors with high sensitivities and to build recognition systems with high correct classification percentages. Many scientists in Japan had pursued their studies on the domain since the 80's of the twentieth century. Nowadays, not only have the basic tastes of sourness, sweetness, bitterness, umami and saltiness been successfully measured, but the quantitive sampling analysis for various foods and beverages, such as coffee, tea, mineral waters and rice, are made considerable headway. Encouraging progress in the domain of machine taste and smell sensation has also been made in our country. Since the 90's of the twentieth century, professor Zhou Chun-Guang in Ji-Lin University, who cooperate with researchers in Saitama University in Japan, pursued his studies on the recognition of taste signals, and firstly applied computational intelligence methods to the recognition for the basic taste signals and their mixture; Professor Wang Ping et al. in Zhe-Jiang University engaged on the researchs of artifical taste and smell, nowadays, some of their achievements have been applied to medicinal diagnoses successfully.Based on understanding and analyzing the actual research state, research focuses and development trend in the domain of machine taste and smell sensation, this dissertation focus on modeling methods for recognizing taste signals. Compared with conventional statistics, this dissertation emphasizes particularly on computational intelligence methods and its integration with other subjects, such as Rough Sets Theory, Statistical Learning Theory, and conventional optimization methods. In conclusion, the main achievements of this dissertation include:(1)This dissertation makes a survey about the research on machine taste and smell sensation, including the appearing background, the actual research state, challenging problems and development trend, etc.(2)In this dissertation, relevant computational intelligence theory, including fuzzy inference model, structural design and learning methods of neural network and Evolutionary Computing theory, are summed up.(3)This dissertation proposes a fuzzy neural network model based on an entropy clustering algorithm. The entropy clustering algorithm improved in this dissertation is superior to the conventional C-means clustering algorithm and the subtractive clustering algorithm in both running speed and deciding parameters. The model employs the Gradient Descent optimization algorithm as parameters learning to refine fuzzy if-then rules, and a system parameter is used to adjust a trade-off between the interpretability and the correct classification percentages, so that the model has good interpretability and learning capability.(4)This dissertation proposes a extended class cover problem and develops a greedy algorithm and a hybrid algorithm to solve the problem. Moreover, the greedy algorithm is used to partition fuzzy input space and extract fuzzy IF-THEN rules to construct a fuzzy neural network model where the particle swarm optimization algorithm is used for optimizing system parameters that improved the system's correct classification percentages and robustness.(5)This dissertation proposes a algorithm to partition fuzzy input space and extract fuzzy IF-THEN rules based on Rough Sets theory. Rough Sets theory was firstly proposed by Z.Pawlak in 1982, but won general popularity in the 90's of the twentieth century. As a mathematical tool to deal with problems with uncertainty, Rough Sets theory and conventional tools, such as Fuzzy Sets and statistics, comp...