A suspension system is a mechanism which consist of spring and damping element connected between wheel and car body. The suspension plays an important role to control the vertical dynamics of car body. The performance and characteristics of suspension system mainly depends on ride comfort and stability control of vehicle . A better ride comfort can be achieved by using soft suspension, whereas better stability can be achieved with the help of hard suspension. The design of suspension involves optimization process where the elements are selected between soft and hard suspension. A suspension is normally classified into passive suspension, semi-active suspension and active suspension. Nowadays, lot of research works are going on [2-5] active suspension system because of its ability to operate wide range of frequency and forces. The performance of active suspension system obtained by measuring suspension travel and acceleration of vehicle body. Due to the development of microcontroller and computers [6-8], the real time implementation of active suspension can be done more effectively. The effect of ride comfort on suspension can be measured with the help of body acceleration of vehicle. Similarly, the performance of stability can be measured with the help of suspension travel.
The flowchart of ANFIS procedure is shown in Figure 4. AN FIS distinguishes itself from normal fuzzy logic systems by the adaptive parameters, i.e., both the premise and consequent parameters are adjustable. The most remarkable feature of the ANFIS is its hybrid learning algorithm. The adaptation process of the parameters of the ANFIS is divided into two steps. For the first step of the consequent parameters training, the Least Squares method (LS) is used, because the output of the ANFIS is a linear combination of the consequent parameters. The premise parameters are fixed at this step. After the consequent parameters have been adjusted, the approximation error is back-propagated through every layer to update the premise parameters as the second step. This part of the adaptation procedure is based on the gradient descent principle, which is the same as in the training of the BP neural network. The consequence parameters identified by the LS method are optimal in the sense of least squares under the condition that the premise parameters are fixed.
(22.2 m 3 s 1 ). Parameter a takes the value 0.129524267 and b T = 0.015 kg min 1 m 2 . The heat transfer coefﬁcient is UA = 25 kW K 1 . Finally, we assume that unknown system and sensor dynamics contribute an overall dead time of 0.5 min in both temperature and humidity measurements (i.e., d T = d w = 0.5 min). Also, we assume that no crop was present in the greenhouse at the time of experiment, but the concrete ﬂoor was continuously wetted to simulate a greenhouse with a wet soil surface. Therefore, the results pre- sented here are supposed to apply to a greenhouse with small seedlings, which do not inﬂuence the greenhouse climate. The greenhouse climate control variables consisted of humidiﬁca- tion and forced ventilation. Suppose this study focuses on daytime control under summer conditions, heating was not considered. A ﬁrst simulation experiment has been conducted to demonstrate the ability of the proposed control schemes to provide interacting control and smooth closed-loop response to set point step changes. For the proposed G-ANFIS con- troller, the dual parameters for each controller are obtained using GA through 50 generations by minimizing the mean square errors,
Neuro-adaptive learning techniques provide a method for the fuzzy modeling procedure to learn information about a data set. It computes the membership function parameters that best allow the associated fuzzy inference system to track the given input and output data. A network-type structure similar to that of an artificial neural network can be used to interpret the input and output map so it maps inputs through input membership functions and associated parameters, and then through output membership functions and associated parameters to outputs. The parameters associated with the membership functions changes through the learning process. The computation of these parameters (or their adjustment) is facilitated by a gradient vector. This gradient vector provides a measure of how well the fuzzy inference system is modeling the input/output data for a given set of parameters. When the gradient vector is obtained, any of several optimization routines can be applied in order to adjust the parameters to reduce some error measure. This error measure is usually defined by the sum of the squared difference between actual and desired outputs. ANFIS uses a combination of back propagation procedure and least squares estimation for membership function parameter estimation.
fuzzycontrol for a category of single-input single-output nonlinear networked control systems with network-induced delay and data loss based on adaptive backstepping control approach. Fuzzy logic systems are used to approximate the unknown nonlinear characteristics existing in the system, while Pade approximation is introduced to handle network- induced delay. Data loss occurs intermittently and stochastically in the data transmitting process,which is regarded as the delay in the controller design. In the framework of adaptivefuzzy back stepping technique, a novel state-feedback adaptive controller is constructed to ensure all signals in the resulting closed-loop system to be bounded and the state variables can be regulated to the origin. Finally, two examples are given to show the validity of the proposed results.
Navigation and obstacle avoidance are very important issues for the successful use of an au- tonomous mobile robot in a dynamic and unstructured environment. Mobile robot researchers aim to build an autonomous and intelligent robot which can plan its motion in a dynamic en- vironment. A successful use of an autonomous mobile robot depends on its controller. Mobile robot control is diﬃcult as they are subjected to non-holonomic (non-integrable) kinematic con- straints involving the time derivates of conﬁguration variables  and dynamic constraints. Both analytical like potential ﬁeld method as well as graph-based techniques have been used to solve the navigation problems of robot involving both static and dynamic obstacles. But, all such methods may not be suitable for on-line implementations due to their inherent computational complexity and limitations. Mobile robot researchers have carried out various researches in this direction using various intelligent techniques methods such as fuzzy logic, neural network and genetic algorithm and their diﬀerent hybrids. Because of the non-linear kinematics of the robot, the uncertainty in sensors readings, and unstructured environmental constrains in the control of mobile robot navigation; researchers have found fuzzy logic as one of the best intelligent tech- nique for handling the constraints. However, fuzzy logic needs tuning for optimal performance. Hand tuning is very diﬃcult and time consuming therefore there is need for automation of the tuning process. The process of tuning requires learning brought about by training or adaptation of the robot to adapt to its dynamic environment. The poor learning capability of fuzzy logic is compensated for by hybridizing fuzzy logic with other soft computing techniques with excellent learning features such as neural network. In this paper, we present an adaptiveneuro-fuzzy controller with genetic algorithm learning for the navigation of Khepera mobile robot.
The basic learning rule of adaptive network is back- propagation algorithm where the model parameters are updated by a gradient descent optimization technique. However, due to the slowness and tendency to become trapped in local minima its application, gradient descent optimization technique, is limited. A hybrid learning algorithm, on the other hand is an enhanced version of the back propagation algorithm . It is applied to adapt the premise and consequent parameters to optimize the network . The hybrid learning rule combines the back- propagation gradient descent method and the least squares estimate (LSE) to update the parameters in the adaptive network. Each epoch of the hybrid learning procedure is composed of a forward pass and a backward pass. The forward pass of the learning algorithm stop at nodes at layer 5 and the consequent parameters are identified by least squares method. After identifying the consequent parameters, the functional signals keep going forward until the error measure is calculated. In the backward pass, the error rate, i.e., the derivative of the error measure with respect to each node output propagates backward from the output end toward the input end, and the premise parameters are updated by the gradient descent method. Heuristic rules are used to guarantee fast convergence. The details of the hybrid rule are given by . The activities in each pass are summarized in Table 1 and flow diagram of ANFIS computations are shown in Fig.2.
Fuzzy logic controller for Inverted Pendulum: It can be built into anything from small, hand-held products to large computerized processcontrol systems. It uses an imprecise but very descriptive language to deal with input data more like a
The neurofuzzy network used in the structure of the proposed hybrid filter acts like a mixture operator and attempts to construct an enhanced output image by combining the information from the new tri-state median (NTSM) filter. The rules of mixture are represented by the rules in the rule base of the neuro-fuzzy network and the mixture process is implemented by the fuzzy inference mechanism of the neuro- fuzzy network. These are described in detail later in this subsection. The neuro-fuzzy network is a first order Sugeno type fuzzy system  with one input and one output. In neuro-fuzzy network, there are two types of fuzzy inference systems are widely used. Mamdani method is widely accepted for capturing expert knowledge. It allows us to describe the expertise in more intuitive, more human-like manner. However, mamdani-type fuzzy inference entails a substantial computational burden. On the other hand, the Sugeno method is computationally effective and works well with optimization and adaptive techniques, which makes it very attractive in control problems, particularly for dynamic nonlinear systems. Sugeno-type fuzzy systems are popular general nonlinear modeling tools because they are very suitable for tuning by optimization and they employ polynomial type output membership functions, which greatly simplifies defuzzification process. The input-output relationship of the neuro-fuzzy network is as follows. Let A 1 denote the inputs of the neuro-
The paper includes the prediction and control of engine air-fuel ratio. Modeling is done by fuzzy clustering and ANFIS. Firstly, inputs and outputs factors of a gasoline engine are replaced as part of system. Later, these factors are grouped into optimal numbers independently by using fuzzy clustering algorithm as a preprocessing step. Later on, these optimal numbers of clustered parameters are used as inputs and outputs of ANFIS for the prediction and controlprocess. Inputs of the system are Manifold Air Pressure (MAP), Throttle Position (TPS), Manifold Air Temperature (MAT), Engine Temperature (CLT), Engine Speed (RPM), and Injection Opening Time (PW) whereas output is AFR, as shown in Figure 2.
They are applied to important fields such as variable speed drives, control systems, signal processing, and sys- tem modeling. Artificial Intelligent systems, means those systems that are capable of imitating the human reasoning process as well as handling quantitative and qualitative knowledge. It is well known that the intelligent systems, which can provide human like expertise such as domain knowledge, uncertain reasoning, and adaptation to a noisy and time-varying environment, are important in tackling practical computing problems. ANFIS has gain a lot of interest over the last few years as a powerful technique to solve many real world problems. Compared to conven- tional techniques, they own the capability of solving prob- lems that do not have algorithmic solution. Neural net- works and fuzzy logic technique are quite different, and yet with unique capabilities useful in information process- ing by specifying mathematical relationships among nu- merous variables in a complex system, performing map- pings with degree of imprecision, control of nonlinear system to a degree not possible with conventional linear systems [5-11]. To overcome the drawbacks of Neural networks and fuzzy logic, AdaptiveNeuro-Fuzzy Infer- ence System (ANFIS) was proposed in this paper. The ANFIS is, from the topology point of view, an implemen- tation of a representative fuzzy inference system using a Back Propagation neural network structure.
Real-time experiment configuration consists of computer with MATLAB, Simulink and Quanser Toolbox used as a controller, Q8 data acquisition board and Quanser IP02 Linear Motion Servo Module. Some hardware limitations should be concerned in the cart- pendulum system. The Digital-to-Analog voltage for data acquisition board is limited between -10 V and 10 V. The safety watchdog is turned on where the allowable cart displacement is 0.35 m from the centre of the track. When the pendulum or cart touches the limit switch, the controlprocess is aborted. Figures 11 to 13 show the SESIP control system experimental results.
Furthermore, the parameter requirement of the varying reluctance machine controller is determined. So the process consists a standardized recurrence of the switch just as inspection with low recurrence. In , a new flexible globally SM control approach based on a global SM control framework and a versatile tracker is implemented to gain predominance after execution control of unsafe and nonlinear time-fluctuating frameworks. Given the normal and exponential term of the primary request, a flexible variable-evaluated exponential methodology law is implemented that has the ability to adjust the sliding surface and the shift in the framework state. The presented method can normally regulate the velocity and weaken the chattering of the structure. Furthermore, on the assumption of the novel arriving at law, AESM control is suggested for BIM velocity control. In order to enhance the execution of sliding mode control against obstruction, an eyewitness aggravation SM is constructed, and its outcome is utilised as AESM control feed forward compensation.
metrological data and (2) Make fuzzy and neurofuzzy model and their results will be evaluated. The fuzzy rule- based approach is applied for the construction of fuzzy models. To remove the weaknesses of fuzzy models that are not trained during the modeling, adaptiveneuro-fuzzy inference system (ANFIS) with given input/output data sets will be use for neuro-fuzzy model. To do this, Fuzzy models was studied with input parameters such as daily maximum and minimum temperature, relative humidity percent, sunshine hours, wind speed. As well as output of this fuzzy system such as Evapotranspiration will studied in this research. After determining effective parameters in
ANFIS is the fuzzy logic based paradigm that grasps the learning abilities of ANN to enhance the intelligent system’s performance using a priori Knowledge. Using a given input/output data set, ANFIS constructs a fuzzy inference system (FIS) whose membership function parameters are tuned (adjusted) using either a backpropagation algorithm alone, or in combination with a least squares type of method. This allows the fuzzy systems to learn from the modeled data. The parameters associated with the membership functions will change through the learning process. The computation of these parameters (or their adjustment) is facilitated by a gradient vector, which provides a measure of how well the fuzzy inference system is modeling the input/output data for a given set of parameters. Once the gradient vector is obtained, any of several optimization routines could be applied to adjust parameters that will reduce some error measure (usually defined by the sum of the squared differences between actual and desired response).
AdaptiveNeuroFuzzy Inference controller (ANFIS) to optimize the performances of photovoltaic techniques. The method consists of a PV panel, a DC–DC booster converter, a maximum power point tracker controller and a resistive load. ANFIS methodology comprises of a hybrid system of fuzzy logic and neural network technique. The fuzzy logic takes into account the decision making based on the rules and uncertainty of the system that is being modeled while the neural network gives it a sense of adaptability. The important thing of the proposed strategy is the use of a ANFIS controller is trained to generate maximum power corresponding to the given solar irradiance level and temperature. The performances of the ANFIS are compared with those obtained using a conventional Adaptivefuzzy controllers with different gains and in each case, the proposed ANFIS controller outperforms its conventional counterpart.
where wn is the output of layer three and [pi and rj] is the i node parameter set. Finally, every node in layer five sums all the incoming signals so that a weighted sum defuzzification technique is performed. The parameter sets o f the FLC antecedents and consequents are tuned or learned using the BP learning algorithm. In [Jang, 1992], the same FFNN configuration was employed but the learning algorithm was a hybrid algorithm. This learning algorithm combined both the BP and the least-square estimation algorithm. In [Yaochu et al., 1995], two interconnected NN were employed to represent a TS-model based FLC. One network represents the antecedent part and the other represents the consequent linear equation. The two NN are then connected through n or product neurons. The BP learning algorithm was employed for learning the parameters o f the consequents and antecedents. Using a TS-model based FFNN has the advantage that it allows a relatively easy mathematical design and stability analysis [Wang and Langari, 1996]. Also, it allows a straight forward application of powerful learning algorithms such as BP due to its differentiable inference functions. On the other hand, a disadvantage o f this model is that the interpretation o f the fuzzy linear rules is difficult compared to that for linguistic rules. Also, the rule base o f this model can only be constructed using only numerical input/output data and it is not possible to incorporate linguistic information from human experts to construct such a model.
control, enhancement of transient stability, mitigation of system oscillations and voltage regulation . A comprehensive and systematic approach for mathematical modeling of UPFC for steady-state and small signal (linearized) dynamic studies has been proposed in , ,  and . The other modified linearized Heffron-Philips model of a power system installed with UPFC is presented in . For systems which are without power system stabilizer (PSS), excellent damping can be achieved via proper controller design for UPFC parameters. By designing a suitable UPFC controller, an effective damping can be achieved. It is usual that Heffron-Philips model is used in power system to study small signal stability. This model has been used for many years providing reliable results .
The responce of the system when we use ANFIS control is better than FLC control but with an overshooting in the dynamic response. In the both system of controls produce a maximum power point voltage of 20.17V as shown in Fig. 10a and Fig. 11b corresponding to characteristic P-V and I-V. The outputs of FLC and ANFIS regulators are connected to the boost converter, so they produce a duty cycle as shown in Fig. 13d. At steady state conditions output power reached the value of 112 W.