61
C H A P T E R 5
Conclusions
is book introduces the design optimization, state estimation, and advanced control methods
for CPVS and their applications in real-world automotive systems.
In Chapter 2, a CPS-based framework for co-design optimization of an automated elec-
tric vehicle with different driving styles was proposed. e multi-objective optimization problem
was formulated. e driving style recognition algorithm was developed using unsupervised ma-
chine learning and validated via vehicle testing. e system modeling and experimental verifica-
tion were carried out. Vehicle control algorithms were synthesized for three typical driving styles
with different protocol selections. e performance exploration methodology and algorithms
were proposed. Test results show that the overall performances of the vehicle were significantly
improved by the proposed co-design optimization approach. Future work will be focused on real
vehicle application of the proposed methods and CPS design methodology improvement.
In Chapter 3, a novel probabilistic estimation method of brake pressure is developed for a
safety critical CPVS based on multilayer ANN with LMBP training algorithm. e high-level
architecture of the proposed multilayer ANN for brake pressure estimation is illustrated at first.
en, the standard BP algorithm used for training of FFNN is introduced. Based on the basic
concept of BP, a more efficient algorithm of LMBP method is developed for model training.
e real vehicle testing is carried out on a chassis dynamometer under NEDC driving cycles.
Experimental data of the vehicle and powertrain systems is collected, and feature vectors for
FFNN training collection are selected. With the vehicle data obtained, the developed multi-
layer ANN is trained. e experimental results show that the developed ANN model, which
is trained by LMBP, can accurately estimate the brake pressure, and its performance is advan-
tageous over other learning-based methods with respect to estimation accuracy, demonstrating
the feasibility and effectiveness of the proposed algorithm. Further work can be carried out in
the following areas: the proposed algorithm will be further refined with onboard road testing;
intelligent control algorithms of braking system will be designed based on state estimation.
High-precision control of mechatronics is an important foundation of the development of
safe, smart and sustainable CPVS [99–104]. In Chapter 4, a typical safety-critical CPVS, i.e.,
the BBW system, was introduced. Compared to the existing BBW system, the newly developed
system enjoys the advantage of a simple structure and low cost because only conventional valves
and sensors are added to the usual hydraulic layouts. Two pressure modulation methods, namely,
the HPBPM and CLPDL modulation, were proposed to improve the modulation precision of
hydraulic brake pressure and reduce valve’s operation noise as well. Experiments were conducted
62 5. CONCLUSIONS
in HiL test rig to demonstrate the performance of the proposed control methods. Experimental
results showed that, in spite of the reduction in noise in the HPBPM control method, when
compared to the conventional PWM control, control accuracy is not guaranteed. e CLPDL
method achieves both good performance of tracking target pressure and reduced noise. To fur-
ther validate the feasibility of the newly proposed BBW system and the CLPDL control method,
a typical regenerative braking scenario was utilized. e CLPDL control method was imple-
mented within a regenerative braking strategy. e HiL test results illustrated that both the
front and rear wheel cylinders fulfilled the driver braking intention and that the regenerative
braking and hydraulic braking cooperate well with each other. For the future work, real vehicle
tests will be conducted, and qualitative comparisons between proposed BBW system and existed
BBW systems will be explored.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.