人工智能在交通指挥中的应用研究

对人工智能进行了研究，分析了机器视觉在道路监视中的作用，指出了监视与控制之间神经网络信号传输中存在的问题．提出了用模式识别技术分解图像中的静态与动态分量的方法。研究表明，采用该方法后，动态数据的传输率可降到1kbit/s。这是一般的无线通信方式都能承受的速率，因此可用窄带无线方法传输图像数据。利用动态数据对道路口各方向运动车辆数量的计算和分析，可预测路口的拥塞程度，实现对路口的智能控制。通过动态数据的图像合成技术实现对监视现场情景的还原，有利于实现交通的人工干预及指挥调度。

Application Research of Artificial Intelligence in Traffic Superintendence

By researching artificial intelligence,the function of machine vision on the supervised road is analyzed, and then, the problem about neuron net signal travel between supervision and control is pointed out, the paper put forth a modes identifying method by which the statical and dynamic component in a picture are divided. The research shows that the rate of the dynamic data could be decreased down to 1kbit/s, it is so low that general radio communication methods can accept it, and the image signal may be traveled by low rate radio. Using the traveled dynamic data, a vehicle number of all direction at traffic across could be calculated and analyzed, and a number of jammed degrees could be forecasted, an intelligence control could be realized. Furthermore, with the help of technology of synthesizing picture to revert the scene of the supervised road, a traffic police could manipulate intelligence system to realize manual interference.