This paper reports on progress made in the first 3 years of.ATR's 'CAM-Brain'Project, which aims to use 'evolutionary e.gi...,i.gi' techniques to build/grow/evolve a RAM-and-cellular-automata based...This paper reports on progress made in the first 3 years of.ATR's 'CAM-Brain'Project, which aims to use 'evolutionary e.gi...,i.gi' techniques to build/grow/evolve a RAM-and-cellular-automata based artificial brain consisting of thousands of interconnected neural network modules inside special hardware such as MITs Cellular Automata Machine 'CAM-8,i, or NTT's Content Addressable Memory System 'CAM-System'. The states of a billion (later a trillion) 3D cellular automata cells, and edlions of cellular automata rules which govern their state changes, can be stored relatively cheaply in giga(tera)bytes of RAM. After 3 years work, the CA rules are almost ready. MITt,,'CAM-8' (essentially a serial device) can update 200,000,000 CA cells a second. It is possible that NTT's 'CAM-System' (essentially a massively parallel device) may be able to update a trillion CA cells a second. Hence all the ingredients will soon be ready to create a revolutionary new technology which will allow thousands of evolved neural network modules to be assembled into artificial brains. This in turn will probably create not only a new research field, but hopefully a whole new industry,namely 'brain building'. Building artificial brains with a billion neurons is the aim of ATR's 8 year i,CAM-B,ai.,' research project, ending in 2001.展开更多
This paper presents two optimization methods for solving the passenger train timetabling problem to minimize the total delay time in the single track railway networks. The goal of the train timetable problem is to det...This paper presents two optimization methods for solving the passenger train timetabling problem to minimize the total delay time in the single track railway networks. The goal of the train timetable problem is to determine departure and arrival times to or from each station in order to prevent collisions between trains and effective utilization of resources. The two proposed methods are based on integration of a simulation and an optimization method to simulate train traffic flow and generate near optimal train timetable under realistic con- straints including stops for track maintenance and praying. The first proposed method integrates a cellular automata （CA） simulation model with genetic algorithm optimiza- tion method. In the second proposed approach, a CA simulation model combines with dynamically dimensioned search optimization method. The proposed models are applied to hypothetical case study to demonstrate the merit of them. The Islamic Republic of Iran Railways （IRIR） data and regulations have been used to optimize train timetable. The results show the first method is more effi- cient than the second method to obtain near optimal train timetabling.展开更多
Landscape metrics are measurements of land- use patterns and land-use change, but even so, have rarely been integrated into land-use change simulation models. This paper proposes a new artificial neural network- cellu...Landscape metrics are measurements of land- use patterns and land-use change, but even so, have rarely been integrated into land-use change simulation models. This paper proposes a new artificial neural network- cellular automaton by integrating landscape metrics into the model. In this model, each cell acquires unique landscape metric values. The landscape metric values of each cell are actually the landscape metric values of land use type in its neighborhood, which takes the cell as center. The calculation of landscape metrics ensures that those of each cell can represent cellular spatial environmental characteristics. The model is used to simulate land use change in the Changping district of Beijing, China. Comparisons of the simulated land use map with the actual map show that the proposed model is effective for land use change simulation. The validation is further carried out by comparing the simulated land use map with that simulated by an artificial neural network-cellular automaton model, which has not been integrated with landscape metrics. Results indicate that the proposed model is more appropriate for simulating both quantity and spatial distribution of land use change in the study area.展开更多
文摘This paper reports on progress made in the first 3 years of.ATR's 'CAM-Brain'Project, which aims to use 'evolutionary e.gi...,i.gi' techniques to build/grow/evolve a RAM-and-cellular-automata based artificial brain consisting of thousands of interconnected neural network modules inside special hardware such as MITs Cellular Automata Machine 'CAM-8,i, or NTT's Content Addressable Memory System 'CAM-System'. The states of a billion (later a trillion) 3D cellular automata cells, and edlions of cellular automata rules which govern their state changes, can be stored relatively cheaply in giga(tera)bytes of RAM. After 3 years work, the CA rules are almost ready. MITt,,'CAM-8' (essentially a serial device) can update 200,000,000 CA cells a second. It is possible that NTT's 'CAM-System' (essentially a massively parallel device) may be able to update a trillion CA cells a second. Hence all the ingredients will soon be ready to create a revolutionary new technology which will allow thousands of evolved neural network modules to be assembled into artificial brains. This in turn will probably create not only a new research field, but hopefully a whole new industry,namely 'brain building'. Building artificial brains with a billion neurons is the aim of ATR's 8 year i,CAM-B,ai.,' research project, ending in 2001.
文摘This paper presents two optimization methods for solving the passenger train timetabling problem to minimize the total delay time in the single track railway networks. The goal of the train timetable problem is to determine departure and arrival times to or from each station in order to prevent collisions between trains and effective utilization of resources. The two proposed methods are based on integration of a simulation and an optimization method to simulate train traffic flow and generate near optimal train timetable under realistic con- straints including stops for track maintenance and praying. The first proposed method integrates a cellular automata （CA） simulation model with genetic algorithm optimiza- tion method. In the second proposed approach, a CA simulation model combines with dynamically dimensioned search optimization method. The proposed models are applied to hypothetical case study to demonstrate the merit of them. The Islamic Republic of Iran Railways （IRIR） data and regulations have been used to optimize train timetable. The results show the first method is more effi- cient than the second method to obtain near optimal train timetabling.
基金This study was supported by the China Postdoctoral Science Foundation （No. 2014M560120）.
文摘Landscape metrics are measurements of land- use patterns and land-use change, but even so, have rarely been integrated into land-use change simulation models. This paper proposes a new artificial neural network- cellular automaton by integrating landscape metrics into the model. In this model, each cell acquires unique landscape metric values. The landscape metric values of each cell are actually the landscape metric values of land use type in its neighborhood, which takes the cell as center. The calculation of landscape metrics ensures that those of each cell can represent cellular spatial environmental characteristics. The model is used to simulate land use change in the Changping district of Beijing, China. Comparisons of the simulated land use map with the actual map show that the proposed model is effective for land use change simulation. The validation is further carried out by comparing the simulated land use map with that simulated by an artificial neural network-cellular automaton model, which has not been integrated with landscape metrics. Results indicate that the proposed model is more appropriate for simulating both quantity and spatial distribution of land use change in the study area.
