Automotive Science and Engineering

---

This paper presents the prediction of vehicle's velocity time series using neural networks. For this purpose, driving data is firstly collected in real world traffic conditions in the city of Tehran using advance vehicle location devices installed on private cars. A multi-layer perceptron network is then designed for driving time series forecasting. In addition, the results of this study are compared with the auto regressive (AR) method. The least root mean square error (RMSE) and median absolute percentage error (MDAPE) are utilized as two criteria for evaluation of predictions accuracy. The results demonstrate the effectiveness of the proposed approach for prediction of driving data time series.

---

In this study the performance and emissions characteristics of a heavy-duty, direct injection, Compression ignition (CI) engine which is specialized in agriculture, have been investigated experimentally. For this aim, the influence of injection timing, load, engine speed on power, brake specific fuel consumption (BSFC), peak pressure (PP), nitrogen oxides (NOx), carbon dioxide (CO2), Carbon monoxide (CO), hydrocarbon (HC) and Soot emissions has been considered. The tests were performed at various injection timings, loads and speeds. It is used artificial neural network (ANN) for predicting and modeling the engine performance and emission. Multi-objective optimization with respect to engine emissions level and engine power was used in order to deter mine the optimum load, speed and injection timing. For this goal, a fast and elitist non-dominated sorting genetic algorithm II (NSGA II) was applied to obtain maximum engine power with minimum total exhaust emissions as a two objective functions.

---

Car following process is time-varying in essence, due to the involvement of human actions. This paper develops an adaptive technique for car following modeling in a traffic flow. The proposed technique includes an online fuzzy neural network (OFNN) which is able to adapt its rule-consequent parameters to the time-varying processes. The proposed OFNN is first trained by an growing binary tree learning algorithm in offline mode, which produces favorable extrapolation performance, and then, is adapted to the stream of car following data, e.g. velocity and acceleration of the target vehicle, using an adaptive least squares estimation. The proposed approach is validated by means of real-world car following data sets. Simulation results confirm the satisfactory performance of the OFNN for adaptive car following modeling application.

---

---

Nowadays, due to increasing the complexity of IC engines, calibration task becomes more severe and the need to use surrogate models for investigating of the engine behavior arises. Accordingly, many black box modeling approaches have been used in this context among which network based models are of the most powerful approaches thanks to their flexible structures. In this paper four network based modeling methods are used and compared to model the behavior of an IC engine: neural networks model (NN), group method of data handling model (GMDH), a hybrid NN and GMDH model (NN-GMDH), and a GMDH model whose structure is determined by genetic algorithm (Genetic-GMDH). The inputs are engine speed, throttle angle, and intake valve opening and closing timing, and the output is the engine brake torque. Results show that NN has the best prediction capability and Genetic-GMDH model has the most flexible and simplest structure and relatively good prediction ability.

.

---

The designing of advanced driver assistance systems and autonomous vehicles needs measurement of dynamical variations of vehicle, such as acceleration, velocity and yaw rate. Designed adaptive controllers to control lateral and longitudinal vehicle dynamics are based on the measured variables. Inertial MEMS-based sensors have some benefits including low price and low consumption that make them suitable choices to use in vehicle navigation problems. However, these sensors have some deterministic and stochastic error sources. These errors could diverge sensor outputs from the real values. Therefore, calibration of the inertial sensors is one of the most important processes that should be done in order to have the exact model of dynamical behaviors of the vehicle. In this paper, a new method, based on artificial neural network, is presented for the calibration of an inertial accelerometer applied in the vehicle navigation. Levenberg-Marquardt algorithm is used to train the designed neural network. This method has been tested in real driving scenarios and results show that the presented method reduces the root mean square error of the measured acceleration up to 96%. The presented method can be used in managing the traffic flow and designing collision avoidance systems.

---

In the present work, the energy absorption study of warm-rolled LZ71 sheet is done for the first time. To do so, Lithium (7% Wt), Zinc (1% Wt) and Magnesium are cast in 770⁰C. After that, the billet has been warm-rolled at 350⁰C and its thickness reduced by 80%. Then, two different heat treatment situations are studied to reach an isotropic plate. Afterward, microstructures of the specimens have been studied using an optical microscope. Tensile tests of the samples are derived to study the mechanical properties and isotropy of the sheets. Moreover, the results of tensile tests applied for crushing simulations. Energy absorption study of the alloy is also done using ABAQUS/Explicit commercial code. The results of simulations are validated using experimental tests of A6082 and completely acceptable performance of simulations is observed. Then, the mechanical properties of LZ71 are used to study the crashworthiness behavior of the mentioned alloy. Crash absorption parameters, namely peak crush force (F_Max), mean crush force (F_Mean), Total Energy Absorption (TAE), Crush Force Efficiency (CFE), Specific Energy Absorption (SEA) and Total Efficiency (TE) of LZ71 and A6082 are compared which are shown that the performance of LZ71 is considerably more efficient than A6082. Lastly, by the help of Artificial Neural Network (ANN) and Taguchi Method, the effects of dimensional parameters of tube, namely diameter, length and thickness, on F_Max, F_Mean and TAE and also the influences of dimensionless geometrical ratios, namely L/D and D/t on CFE, SEA and TE are surveyed comprehensively.

---

In this paper, an adaptive cruise control system is designed that is controlled by a neural network model. This neural network model is trained with data resulting from the simulation of a multi-objective nonlinear predictive adaptive cruise control system. For this purpose, first, an adaptive cruise control system was designed using the concept of model predictive control based on a nonlinear model to maintain the desired speed of the driver, maintain a safe distance with the car in front, reducing fuel consumption and increasing ride comfort. Due to the time-consuming computations in predictive control systems and the consequent need for powerful and expensive hardware, it was decided to use the extracted data from the simulation of this designed cruise control system to train a neural network model and use this model to achieve control objectives instead of the predictive controller. Using the neural network model in the cruise control system, despite a significant reduction in computation time, the control objectives were well achieved, and in fact a combination of model predictive controller accuracy and neural network controller speed was used.