---

Considering that once-through Benson boiler is one of the most crucial equipment of a thermal power plant, occurrence of any fault in its different parts can lead to decrease of the performance of system, and even may cause system damage and endanger the human life. In this paper, due to the high complexity of the system's dynamic equations, we utilized data-based method for diagnosing the faults of the once-through Benson boiler. In order to enhance the fault diagnose (FD) system proficiency and also due to strong interactions between measurements, we decided to utilize six one-class support vector machine (SVM) algorithms to diagnose six major faults of once-through Benson boiler. In the proposed structure, each One-class SVM algorithm has been developed to diagnose one special fault. Finally, we carry out diverse test scenarios in different states of fault occurrence to evaluate the performance of the proposed FD system against the six major faults of the once-through Benson Boiler under conditions of noisy measurement.

 
 

---

Nowadays, a great number of researches are done by scientists to provide some models that can predict the passenger injuries in crashes. In this paper, a hybrid model of vehicle and passenger is proposed to predict the head acceleration in the front crash. A lumped mass model with 12-degree-of-freedom (DOF) is firstly used to predict the behavior of vehicle in front crash. In this model, any member of vehicle is modeled as a lumped mass and connected to the other members through some springs and dampers. The unknown coefficients of such model are obtained using genetic algorithm to minimize the deviation between the results of experimental and suggested model. The parameters of model are established by experimental results of a real world car, namely the HONDA ACORD2011, in an accident velocity of 48 km/h. Also, the validity of the proposed model is checked by experimental results of mentioned vehicle at two other crash velocities of 40 km/h, and 56 km/h. The results show that the proposed model is an efficient framework for preliminary designing of both structure and parameter design of vehicle to improve its crash worthiness. Moreover, a multi-body dynamic model of driver is proposed to predict the head injury in front crash. The seat acceleration which has been calculated using vehicle’s model is considered as input of this model.

---

In this study, an application of support vector machines are presented for fouling detection and estimating the amount of deposit layer development and tube blockage percent at the radiation section of the crude oil preheat furnace. Crude oil preheat furnaces are the main elements in processing crude oil in distillation towers, which may always suffer from fouling and its consequent risks. In order to predict fouling inside the tubes, first by considering independent input parameters effecting the furnace performance and by using a dynamic model of a particular furnace, the behaviors of the furnace in unusual conditions were simulated. The effects of fouling type and its location inside the tubes were considered on the thermal performances and pressure drops of the furnace. In the second part, based on the different fouling scenarios, a fouling detection mechanism was designed. The operational conditions such as pressure drop inside the tubes, temperatures of the tubes and temperatures of the crude oil were employed for fouling detection and evaluating the thickness of deposits. The obtained results indicated the accuracy and feasibility of proposed approach.

---

In this study, feedback-feedforward control system design and optimizing the performance of crude oil furnace process was investigated in order to be recovered from possible abnormal conditions. First, by developing an accurate nonlinear analytical model, the effects of changes in input parameters and operating conditions on the system’s outputs were determined. Then, in order to eliminate the effects of disturbances on furnace, a feedback- feedforward control system for combustion management was suggested, where its performances were optimized genetic algorithm (GA). In addition, to enhance the thermal stability and to maintain product quality, output difference temperature control system was considered for load distribution between furnace’s streams. Also, in order to recover the furnace from abnormal conditions due to burners’ failures, a supervisory system was designed to change the firing rate setpoints. With respect to different failure scenarios, the optimal burners’ firing rate were captured by applying genetic algorithms to the system model. A multilayer perceptron neural network was employed as the core of the controller to interpolate between different conditions. The obtained results indicate the superior performances of the designed control systems.

---

In this study, an application of support vector machine (SVM) for early fault detection in increasing the level of the start-up vessel in a Benson type once-through boiler during load changes is presented. The level increasing in the start-up vessel is happened due to thermal conditions disruption inside the boiler especially while the unit load is ramped-down. In this regard, first, the variables effective on increasing the level of start-up vessel was identified based on experimental data from a power plant unit, then the dimension of input variables was reduced by selecting appropriate features. Experimental results show that the hotwell surfaces’ temperature could be considered as the most appropriate indicator for steam quality deterioration. By comparing the extracted features from healthy and unhealthy conditions, appropriate fault model was developed using SVM with radial basis function (RBF) as the kernel. The performances of fault detection system were evaluated with respect to the similar faults at two different time periods happen in a steam power plant. The obtained results show the accuracy and feasibility of the proposed approach in early detection of faults during the unit’s load variations. Advantages of the proposed technique is preventing false alarm in power plants’ boilers as load changes.

---

In this study, fair comparisons between the empirical mode decomposition, ensemble empirical mode decomposition and discrete wavelet transform with the mother wavelet function of Meyer and Daubechies, were performed for detecting unbalance faults in a rotating machinery. In order to classify the healthy class from the unbalance classes, a support vector machines that was optimized by particle swarm optimization algorithm, was used. A comparison between the performances of optimized and non-optimized of support vector machines were also carried out. In order to obtained the required data, a rotating machinery fault simulator was developed and vibrational signals were acquired at healthy and unbalance fault conditions by accelerometer sensors. By processing the recorded signals and analysing signal to their frequency components, several statistical features were extracted from each frequency component as input support vector machine for the separation of classes. The obtained results indicated that the discrete wavelet transform with the Meyer mother wavelet, higher success rate than other methods for diagnosing unbalance faults.