Simulation Of Fire Prediction And Research On Path Planning

Fire has been a topic of concern, fire prevention is also the research topic in these years. Especially in recent years. The city changes gradually speed up the process, city is becoming more and more crowded, high-rise is bristly, population density, a variety of fire factor is with them, these are the hidden trouble of fire. At the same time, as the fire occurs, the life and property of people often face serious threat. In the previous example of fire, a few examples caused the death, and a large part of the reason is the emergency lights. Obviously a good emergency indicating lights design plays an important role in the process of guiding people to escape.In order to establish a good fire early warning system and a good emergency indicator system. This paper will focus on the fire information process and path planning of indicator light.In the fire information processing, this paper adopts fuzzy neural network algorithm to process the fire information, abandoned the previous way of fire early warning only by hardware, strive to improve the efficiency of early warning of fire. In information processing, using the method of neural networks and fuzzy control for connecting. To start with, fire detecting is not a single signal detection, instead of using the composite signal detection. According to the fire characteristics, this paper will detect three fire factor, including CO, Smoke, temperature. Secondly, using the BP neural network processes the detected information and make a comprehensive comparison. It is helpful to improve the efficiency of early warning. Thirdly, use fuzzy algorithm to process the result which has been processed by BP neural network. When setting a threshold, sometimes there will be a dilemma by using fuzzy judgement. If the result of the judgement is to the critical point, it will make the fire warning efficiency decreased. In order to improve the decision efficiency as far as possible, this paper use the fuzzy algorithm to step process the result after the neural network. At last.according to results, we can judge the fire.In the light path planning, this paper will make an automated processing about indicating light, rather than the traditional mechanical model. Firstly, make the requirement of all lights constructed into a whole, in accordance with the present commonly used communication mode, using serial communications, the central processor can control whole indication light system. Thus indication light can follow the directives issued by the central system to complete the action. Followed by the path information for planning, once there is a fire, indicating light must be able to indicate proper safe passage for all staff. For less complex floor, exhaustive method can be used to make assumptions for all circumstances, then work out the safe and correct path. But once the floor layout is so complicated, using exhaustive method is not suitable, computation is very big and often makes a lot of mistakes. In this paper, using Floyd algorithm to complete path planning, it not only reduce the workload, but also improve the accuracy of path planning. Finally, according to the path planning, write procedures to control the indicating light.