目录
摘要 II
关键词 II
Abstract III
引言
引言 1
1 数据处理 1
1．1 定义不良驾驶行为 1
1．1．1 怠速 1
1．1．2 疲劳驾驶 1
1．1．3 急加速 2
1．1．4 急减速 2
1．1．5 熄火滑行 2
1．1．6 急转弯 2
1．1．7 超速 3
1．1．8 停车熄火不拉手刹 3
1．2 定义不良驾驶行为发生率 3
2 基于回归分析的神经网络模型 3
2．1 建立评估项目权重值 4
2．1．1 八项评价指标定义 4
2．1．2 设定评估项目的权重 4
2．2 基于回归分析的神经网络模型 5
2．2．1 回归分析模型 5
2．2．2 神经网络算法 6
3 结果与分析 7
3．1 结果 7
3．1．1 八个不良驾驶行为的发生率 7
3．1．2 求解综合评价值 7
3．1．3 建立一元线性回归模型 8
3．1．4 预测 10
3．1．5 神经网络预测安全等级 11
3．2 分析 13
4 讨论 14
4．1 运输车辆信息包含天气等分类数据 14
4．2 建立回归模型时输入变量取值未知 14
致谢 14
参考文献 15
附录 16
运输车辆安全驾驶行为分析的数学模型
摘 要
本文将回归分析法与神经网络算法相结合，建立出一套完整的评价运输车辆安全等级的数学模型。首先计算出全部运输车辆的8项不良驾驶行为的发生率，并将运输车辆分为三个小样本群。对第一个样本群，根据箱线图划分等级结果以及运输车辆的综合评价值得到回归函数；对第二个样本群，利用回归函数的预测功能得出其安全评价等级结果，并把第一、二个样本群作为神经网络模型中的 *51今日免费论文网|www.51jrft.com +Q: ￥351916072￥ 
训练集，建立多分类的神经网络预测模型；对第三个样本群利用神经网络模型进行预测，并给出8项不良驾驶行为对评价等级的重要性排序。三个样本群的样本量依次为10、40、50个。不良驾驶行为对评价运输车辆安全等级的影响从大到小依次为：急转弯和疲劳驾驶、急减速、超速、急加速、停车熄火不拉手刹、熄火滑行和怠速预热。模型预测出等级评价为极好、优、良好、一般、差、极差的运输车辆依次占比0%、14%、32%、24%、24%、6%，说明当前车辆行驶情况较乐观。
MATHEMATICAL MODEL OF TRANSPORTATION VEHICLE SAFE DRIVING BEHAVIOR ANALYSIS
ABSTRACT
This paper combines regression analysis and neural network algorithm to establish a complete set of mathematical models for evaluating the safety level of transportation vehicles. First calculate the incidence of 8 bad driving behaviors for all transportation vehicles, and divide the transportation vehicles into three small sample groups. For the first sample group, the regression function is obtained according to the classification results of the box plot and the comprehensive evaluation value of the transportation vehicle; for the second sample group, the prediction function of the regression function is used to obtain the safety evaluation level result 2. The two sample groups are used as the training set in the neural network model to establish a multiclass neural network prediction model; the third sample group is predicted using the neural network model, and the importance of 8 bad driving behaviors on the evaluation level Sort. The sample sizes of the three sample groups are 10, 40 and 50 in order. The influence of bad driving behaviors on the evaluation of the safety level of transportation vehicles from large to small is as follows: sharp cornering and fatigue driving, sharp deceleration, overspeed, rapid acceleration, parking without stopping the handbrake, flameout and idle warmup. The model predicts that the transportation vehicles with grade evaluation of excellent, excellent, good, fair, poor, and poor account for 0%, 14%, 32%, 24%, 24%, and 6% in turn, indicating that the current vehicle driving situation is more optimistic .

原文链接：http://www.jxszl.com/jsj/sxtj/606749.html