Research On Date Processing Algorithm Of Dynamic Weighting System

The excess phenomenon that the local highway transport is very serious, the excess is the main reason of the highway damage, the dynamic state ground diagraph drive medium the stalk of vehicle carry, for protect of highway and enforce the law meaning of have the importance.Weigh system but speech for the highway dynamic state, dynamic state's measuring pass of accuracy, car speed is the most important in its function index sign and it symbolizes the technique that the highway dynamic state weighs system Gao Di of the level.The highway dynamic state weighs system structure style a lot of, the different structure comes to a decision the method dissimilarity that the follow-up weighs a signal processing, but signal treatment be one of the key factors which influences system accuracy.The current dynamic state weighs system because of weighing signal to make simple numeral to filter a reason just to the dynamic state, lack more deeply the signal of the 1 F a processing a technique, so the accuracy of the system is hard to get a very great exaltation.This text to the signal processing part that the dynamic state weighs system carries on a research, main research the work include a few aspects as follows:1)The dynamic state weighs a signal frequency chart analysis and prepares a processing.Pass to weigh the analysis of factor of the signal interference to the dynamic state, adopt 2 to prepare treatment:The numerical low of the square one adoption filters a method and filter in addition to the interference Zao of the high frequency segment voice;The second step adoption is excellent to turn calculate way to draw up to match the frequency in a signal very low, and mixed with useful signal together of low frequency interference signal, and give cancellation.2)Weigh system model a research according to the vehicle dynamic state of parameter estimate calculate way.Recognize a method mainly knowledge according to the first check of the examination system according to the round heavy dynamic state of parameter estimate, deduce to imply the model which didn't know parameter, then make use of the importation, exportation data of system to carry on system parameter to recognize.Introduced a dynamic state to weigh the ARX mechanism of system model first, and as to it's carried on to imitate a true research.3)Owing to the influence dynamic state weigh of the interference factor be a lot of, and the nonexistent accurate function relation of these factors, use a traditional mathematics method very difficult analytical the relation of of these interference factors.But artificial nerve the network has it the special advantage in the processing not- line and complicated problem.This text aims at the BP network existent blemish and constructed 1 kind to minimum advantage of two multiplications of the nerve network and the line before combine according to mixture structure of nerve network, put forward homologously non- Die generation of fast study calculate way.Should study calculate way can according to draw up to match an accuracy request, make use of line minimum two multiplicationses to make sure homologous best network power value and line part of parameter, and auto certain the best concealed layer node number.Express with the comparison result of BP network, this text put forward of hybrid structure front to nerve network of fast study the calculate way is regardless draw up to match accuracy, studies speed, is suffused with the ability of turn, still concealed node the number all shows the Zhao good to BP network.