Characterizing human driver characteristics using an artificial neural network and a theoretical model

Abstract

Human drivers seem to have different characteristics, so different drivers often yield different results from the same driving mode tests with identical vehicles and same chassis dynamometer. However, drivers with different experiences often yield similar results under the same driving conditions. If the features of human drivers are known, the control inputs to each driver, including warnings, will be customized to optimize each man–machine vehicle system. Therefore, it is crucial to determine how to characterize human drivers quantitatively. This study proposes a method to estimate the parameters of a theoretical model of human drivers. The method uses an artificial neural network (ANN) model and a numerical procedure to interpret the identified ANN models theoretically. Our approach involves the following process. First, we specify each ANN driver model through chassis dynamometer tests performed by each human driver and vehicle. Subsequently, we obtain the parameters of a theoretical driver model using the ANN model for the corresponding driver. Specifically, we simulate the driver’s behaviors using the identified ANN models with controlled inputs. Finally, we estimate the theoretical driver model parameters using the numerical simulation results. A proportional-integral-differential (PID) control model is used as the theoretical model. The results of the parameter estimation indicate that the PID driver model parameter combination can characterize human drivers. Moreover, the results suggest that vehicular factors influence the parameter combinations of human drivers.

Appendices

Appendix A: Experimental history information

$$\begin{aligned}&\text {Error distance (m)}=\int v_{\mathrm{target}}-v_{\mathrm{experiment}}\mathrm{d}t, \end{aligned}$$

(A1)

$$\begin{aligned}&\text {Ratio of error distance}=\frac{\text {Error distance (m)}}{\text {Mode distance (km),}} \end{aligned}$$

(A2)

where $v_{\mathrm{target}}$ is the target speed of the driving modes, $v_{\mathrm{experiment}}$ is the experimental vehicle speed, and mode distance (km) is the total distance of each driving mode (Table 6).

Table 6 Experimental history information

Appendix B: Additional simulation results

Table 7 Selected constants of v, a, $v_{P}$, and $a_{D}$ as input data for identified ANN driver models (1)

We have performed additional $K_{P}$ and $\tau _{D}$ estimations under the $v_{P}$ and $a_{D}$ conditions different from those presented in Tables 7 and 8. The results are illustrated in Figs. 11, 12, 13 and 14. A similar trend was observed in the results, as illustrated in Figs. 9 and 10. Differences in K and $\tau$ were observed depending on the vehicle speed conditions, whereas the trends of $K_{P}$ and $\tau _{D}$ depended on the drivers.

Table 8 Selected constants of v, a, $v_{P}$, and $a_{D}$ as input data for identified ANN driver models (2)

Fig. 11

Estimated K and $\tau$ gains for each driver and vehicle (Condition Table 7). a $V_1$. b $V_2$. c $V_3$ HEV. d $V_3$ EV

All the $K_{P}$ and $\tau _{D}$ estimation results in this paper indicate that the tested drivers have different $K_{P}$–$\tau _{D}$ combinations for the three vehicles. In particular, the results for HEV and EV show that the drivers’ $K_{P}$ and $\tau _{D}$ data are plotted in different regions in the $K_{P}$–$\tau _{D}$ planes for each driver under all the driving conditions, i.e., all the combinations of the values of v, a, $v_{P}$, and $a_{D}$. In addition to that, the $V_1$ drivers’ plotted data regions in the $K_{P}$–$\tau _{D}$ plane show a similar tendency except for the conditions including the smallest $v_{P}$, and $a_{D}$ values. In the case of the smallest $v_{P}$ and $a_{D}$ combination, since the influence of the difference between the current and target values is small, the difference in the drivers’ parameters may not be so significant. On the other hand, the results in terms of $V_2$ show that the tested drivers have almost the same plotted data regions. The vehicle characteristics may cause this.

Fig. 12

Estimated $K_{P}$ and $\tau _{D}$ gains for each driver and vehicle (Condition Table 7). a $V_1$. b $V_2$. c $V_3$ HEV. d $V_3$ EV

Fig. 13

Estimated K and $\tau$ gains for each driver and vehicle (Condition Table 8). a $V_1$. b $V_2$. c $V_3$ HEV. d $V_3$ EV

Fig. 14

Estimated $K_{P}$ and $\tau _{D}$ gains for each driver and vehicle (Condition Table 8). a $V_1$. b $V_2$. c $V_3$ HEV. d $V_3$ EV