-
从安全行车视距模型的建立过程可以看出,车载红外辅助驾驶系统的安全车速与路面条件和探测距离有直接关系,最大安全车速
${V_{\max }}$ 满足式中:滑动附着系数
${\varphi _{\rm s}}$ 与路面条件直接相关[15],常见路面土路和混凝土路的滑动附着系数如表1所示。Pavement type Dry concrete road Wet concrete road Dry dirt road Wet dirt road Muddy dirt road Sliding adhesion coefficient 0.75 0.35 0.65 0.20 0.15 Table 1. Sliding adhesion coefficient of common pavement type
路况一定时,辅助驾驶的最大安全车速主要取决于动态探测概率和红外探测器的探测距离。干燥土路面和干燥混凝土路面条件下,取动态探测概率
${p_2} = 99\% $ ,探测距离对最大安全车速的影响如图2所示。可以明显看出二者的正相关关系,当探测距离分别大于190 m左右、160 m左右时,最大安全车速达到了120 km/h。为了进一步分析环境条件对最大安全车速的影响,需对探测距离
$R$ 进行求解,求解之前首先对$MRTD(f)$ 进行修正。$MRTD(f)$ 的大小一般来自于实验室测试或理论计算,通过车载红外辅助驾驶系统实际观察目标时,由于不满足实验室标准条件和理论计算理想条件,可对一些因子进行修正。以噪声等效温差
$NETD$ 为基础,考虑热成像系统全部电路的带宽及人眼作用,$MRTD(f)$ 可表示为:式中:
$SN{R_{\rm DT}}$ 为阈值信噪比;$MT{F_{\rm S}}\left( f \right)$ 为热成像系统总的调制传递函数;$\alpha $ 、$\beta $ 分别为探测器横向和纵向的角分辨率;${\tau _{\rm d}}$ 为探测器的积分时间;${f_{\rm P}}$ 为帧频;$\Delta {f_{ n}}$ 为放大电路等效噪声带宽。实验室MRTD测试采用的是长宽比为7:1的四条带标准靶,辅助驾驶过程中路面目标的实际高宽比不一定满足理想条件,因此,需要进行目标形状的修正。目标形状修正因子为
${k_1} = \sqrt {7/\varepsilon } $ ,其中,$\varepsilon $ 表示目标等效高宽比,与实际高宽比${\varepsilon _0}$ 的关系可表示为:同时,实验室
$MRTD$ 测试采用的是恒定温度为${T_{\rm T}}$ 的黑体目标背景,辅助驾驶过程中需根据实际路面温度${T_{\rm m}}$ 进行修正,修正因子为${k_2} = {T_{\rm T}}/{T_{\rm m}}$ 。车载红外辅助驾驶系统对路面目标的静态探测概率
${p_1}$ 与阈值信噪比$SNR$ 有关,其关系为:阈值信噪比修正因子为
${k_3} = SNR/SN{R_{\rm DT}}$ 。修正后的
$MRTD(f)$ 可表示为:则有:
可以看出,最大安全车速
${V_{\max }}$ 的影响因素主要包括探测概率${p_1}$ 和${p_2}$ 、目标尺寸、目标与背景等效温差、大气透过率、目标高度、目标等效条带对数。辅助驾驶过程中,为确保行车安全,可取静态探测概率${p_1} = 100\% $ 。目前,车载红外辅助驾驶系统一般采用8~14 μm波段探测器,当探测距离为300 m时,通过LOWTRAN计算不同能见度下的大气透过率,结果如图3所示。可以看出,由于车载红外辅助驾驶系统为近距离探测,天气条件较好时能见度对大气透过率的影响不明显。
可见观察等级和目标一定时,良好天气下的最大安全车速主要与目标背景温差有关。图4为干燥土路面和干燥混凝土路面下,观察等级分别为发现和识别时,最大安全车速与目标背景温差的变化关系,仿真计算过程中基本参数值如表2所示。从图中可以看出,随着温差的增大最大安全车速也明显增大,且观察等级为发现时的最大安全车速远大于识别时。辅助驾驶过程中,为保证行车安全应以识别条件下的车速控制为主,当路况条件较好时可选择发现条件下的车速控制。
SNRDT ${\tau _{\rm{d}}}$ Teye fP NETD H p1 p2 2.8 30 µs 0.1 s 50 Hz 40 20 cm 1 0.99 Table 2. Basic parameter value of simulation calculation
Influences of weather conditions on vehicular infrared assistant driving performance
doi: 10.3788/IRLA20190507
- Received Date: 2019-11-15
- Rev Recd Date: 2019-12-20
- Available Online: 2020-04-30
- Publish Date: 2020-07-23
-
Key words:
- vehicular infrared assistant driving system /
- assistant driving performance /
- maximum safe speed /
- detection distance /
- sliding adhesion coefficient
Abstract: Based on comprehensive analysis of vehicle infrared assisted driving safety braking process, infrared static detection model and dynamic detection model, the model of visual range for assistant driving safety was established. The relationship between maximum safe speed, road condition, infrared detection performance was analyzed. It is found that it is mainly related to sliding adhesion coefficient and detection distance. Through the correction calculation of infrared detection distance, the maximum safe speed at night was simulated. The results show that it is mainly affected by the temperature difference between target and background. Emphasis is laid on foggy and rainy weather for example analysis. The results show that foggy weather mainly affects the detection distance, especially when the visibility is less than 1 km, and the maximum safe speed need be controlled at 21-25 km/h when the visibility is 5 km. Rainy weather will affect the sliding adhesion coefficient and detection distance. The maximum safe speed control under infrared recognition is the main method for assistant driving. When rainfall intensity is 50 mm/h, the maximum safe speed need be controlled at 12-14 km/h.