Research Article

Vehicle Information Influence Degree Screening Method Based on GEP Optimized RBF Neural Network

Table 6

The operating driving range influence factors forecasting results of statistics for pure electric bus (relative error values).

Influence factorPrediction objectSample sizeNumber of sample error less than 5%Proportion of the sample error less than 5%Number of sample error less than 10%Proportion of the sample error less than 10%Number of sample error less than 20%Proportion of the sample error less than 20%Number of sample error more than 20%Proportion of the sample error more than 20%

Minimum temperatureWorkdays799901434417.93%3892048.66%6587682.36%1411417.64%
Weekends22686298813.17%1009844.51%1800479.36%468220.64%
Holidays23294278011.93%987142.38%1823278.27%506221.73%

Energy feedback mileage proportionWorkdays799901502118.78%4098751.24%6721384.03%1277715.97%
Weekends2268617427.68%1084447.80%1598070.44%670629.56%
Holidays2329418928.12%810934.81%1623169.68%706330.32%

Traffic congestion indexWorkdays799901800222.51%41,98452.49%7002487.54%996612.46%
Weekends22686367416.20%1409162.11%2012588.71%256111.29%
Holidays23294440218.90%1299055.77%1703473.13%626026.87%

Sliding time proportionWorkdays799901220815.26%3999550.00%6402980.05%15,96119.95%
Weekends22686301113.27%1509166.52%1922184.73%346515.27%
Holidays23294267711.49%1304456.00%1730274.28%599225.72%

Energy feedback proportionWorkdays799901720121.50%4220152.76%6298178.74%17,00921.26%
Weekends22686231210.19%998144.00%1547268.20%721431.80%
Holidays23294250810.77%1464762.88%1400660.13%928839.87%

Braking mileage proportionWorkdays79990974812.19%3778447.24%5902773.79%2096326.21%
Weekends2268616707.36%1223153.91%1309857.74%958842.26%
Holidays2329417327.44%1435261.61%1500764.42%828735.58%

Uniform mileage proportionWorkdays799901800922.51%4499856.25%6803985.06%1195114.94%
Weekends22686409818.06%1202252.99%1578069.56%690630.44%
Holidays23294311913.39%1300555.83%1642470.51%687029.49%

Note: the “number of sample error less than 10%” containing the “number of sample error less than 5%” and “number less than 5% error value is less than or equal to 10%” can be calculated like “number of sample error less than 10%” minus “number of sample error less than 5%.” Other similar.