Improved Back Propagation Algorithm Research Articles

Research on Improved Quantitative Identification Algorithm in Odor Source Searching Based on Gas Sensor Array.

In order to improve the precision of gas detection and develop valid search strategies, the improved quantitative identification algorithm in odor source searching was researched based on the gas sensor array. The gas sensor array was devised corresponding to the artificial olfactory system, and the one-to-one response mode to the measured gas was set up with its inherent cross-sensitive properties. The quantitative identification algorithms were researched, and the improved Back Propagation algorithm was proposed combining cuckoo algorithm and simulated annealing algorithm. The test results prove that using the improved algorithm to obtain the optimal solution -1 at the 424th iteration of the Schaffer function with 0% error. The gas detection system designed with MATLAB was used to obtain the detected gas concentration information, then the concentration change curve may be achieved. The results show that the gas sensor array can detect the concentration of alcohol and methane in the corresponding concentration detection range and show a good detection performance. The test plan was designed, and the test platform in a simulated environment in the laboratory was found. The concentration prediction of experimental data selected randomly was made by the neural network, and the evaluation indices were defined. The search algorithm and strategy were developed, and the experimental verification was carried out. It is testified that the zigzag searching stage with an initial angle of 45° is with fewer steps, faster searching speed, and a more exact position to discover the highest concentration point.

Optimal parameters selection of back propagation algorithm in the feedforward neural network

The back propagation (BP) is one of the most widely used algorithms in the feedforward neural network (FNN), but selecting the non-optimal weights and thresholds may induce the slow convergence and local optimization. In this work, we propose an improved BP algorithm in which the loss function is constructed based on a small scale of data to seek the optimal weights and thresholds. The iteration process of the improved BP algorithm can be accelerated and the solutions will not be trapped into the local optimum. Solutions corresponding to the increasing scales of data are convergent which demonstrates the convergence of the parameters selection. Numerical simulations of elasticity and graphic reconstruction and repair confirm that the improved BP algorithm can achieve much faster convergence and higher efficiency than the conventional BP algorithm. Moreover, different initial values leading to the same optimal parameters indicates that the improved algorithm can achieve the global optimization during the learning process.

A novel fault diagnosis scheme for a ballast system

The Back Propagation (BP) neural network model is one of the neural network models suitable for model recognition and classification. However, a single BP network model cannot fully tap the feature information hiding in the data and is vulnerable to the uncertainty associated with system parameter changes and also suffers from the effects of long diagnosis times, inaccurate and undesired diagnosis results in applying a large amount of high-dimensional feature data. In addition, the standard BP algorithm has limitations in pattern recognition and classification. For example, it adopts the quadratic cost function to train the neural network, which can lead to low convergence accuracy, low convergence rate, and so on. Compared with the quadratic cost function, the cross-entropy cost function can effectively improve the convergence accuracy and the convergence rate of the neural network. The Dempster–Shafer (D–S) evidence theory is a powerful tool for handling incomplete, imprecise, and uncertain information and merging it when it derives from multiple BP networks. In order to overcome the drawbacks of a single BP network, it is necessary to divide the high-dimensional feature data into several different categories of low-dimensional feature data and to input the same dimensional feature data into the same BP network model, as well as to design the information fusion algorithms based on D–S evidence theory and multiple improved BP networks with the cross-entropy cost function. For this reason, this paper proposes a novel information fusion diagnosis approach based on D–S evidence theory and multiple improved BP networks for fault diagnosis and classification of multiple devices in a ballast system subjected to extreme environmental conditions. The proposed diagnosis approach utilizes multiple neural networks to carry out local fault diagnosis, adopts the D–S evidence theory to fuse the results of local diagnosis and derive the decision results, and introduces an improved BP algorithm based on the cross-entropy cost function into the individual neural networks to enhance the convergence rate and diagnosis accuracy of the individual neural networks. This method can overcome the deficiencies of individual BP networks in multiple BP networks in their poor diagnostic ability to handle incomplete, imprecise, and uncertain sample information. According to the design idea of the proposed diagnosis approach, a new fusion modeling for fault diagnosis of multiple devices is constructed for a ballast system. The validity of the proposed diagnosis method and model is illustrated through both simulation studies and experimental tests.

The Analysis of Differential High-Quality Development of Economy by Deep Neural Network and Internet of Things

This study solves the problems of insufficient objectivity and a large amount of calculation in traditional regional economic development level evaluation methods. The conventional Back Propagation (BP) neural network can easily fall into local extremes and slow convergence speed. Firstly, an improved BP algorithm is proposed. The LM (Levenberg-Marquardt) algorithm optimizes the BP neural network. Secondly, the evaluation model of Henan County’s economic development level is constructed based on the improved BP algorithm. Finally, Matlab software is used to design simulation experiments. The BP model is used to evaluate the comprehensive economic development level of 107 counties in Henan Province. The results show that the Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Mean Square Error (MSE) of the output of the proposed BP model are 0.9, 1.72, and 3, respectively. This is far lower than other popular algorithms and is more suitable for estimating the comprehensive development level of the region. The overall economic development level of counties in Henan Province is relatively balanced but partially uneven. The polarization of development is more serious. The middle-level and middle- and low-level development counties account for a relatively large area, and the overall distribution characteristics are “convex”. This study aims to provide necessary technical support for a more accurate analysis of Henan counties’ comprehensive economic development level distribution and then to promote the coordinated development of counties in the province.

Flexible Transmitter Network

Current neural networks are mostly built on the MP model, which usually formulates the neuron as executing an activation function on the real-valued weighted aggregation of signals received from other neurons. This letter proposes the flexible transmitter (FT) model, a novel bio-plausible neuron model with flexible synaptic plasticity. The FT model employs a pair of parameters to model the neurotransmitters between neurons and puts up a neuron-exclusive variable to record the regulated neurotrophin density. Thus, the FT model can be formulated as a two-variable, two-valued function, taking the commonly used MP neuron model as its particular case. This modeling manner makes the FT model biologically more realistic and capable of handling complicated data, even spatiotemporal data. To exhibit its power and potential, we present the flexible transmitter network (FTNet), which is built on the most common fully connected feedforward architecture taking the FT model as the basic building block. FTNet allows gradient calculation and can be implemented by an improved backpropagation algorithm in the complex-valued domain. Experiments on a broad range of tasks show that FTNet has power and potential in processing spatiotemporal data. This study provides an alternative basic building block in neural networks and exhibits the feasibility of developing artificial neural networks with neuronal plasticity.

Analysis of artificial neural network performance based on influencing factors for temperature forecasting applications

Artificial neural network (ANN)-based methods belong to one of the most growing research fields within the artificial intelligence ecosystem, and many novel contributions have been developed over the last years. They are applied in many contexts, although some “influencing factors” such as the number of neurons, the number of hidden layers, and the learning rate can impact the performance of the resulting artificial neural network-based applications. This paper provides a deep analysis about artificial neural network performance based on such factors for real-world temperature forecasting applications. An improved back propagation algorithm for such applications is also presented. By using the results of this paper, researchers and practitioners can analyse the encountered issues when applying ANN-based models for their own specific applications with the aim of achieving better performance indexes.

PSD Integrated Calibration Method Based on Annunciator in Vacuum Environment

In this paper, an integrated calibration method for position sensitive detector (PSD) of moving support mechanism in vacuum plasma environment is proposed. That is, the PSD is used to track the laser spot to restore the coordinate information of the signal. Firstly, a PSD integrated calibration principle is proposed to reduce the position error of the signaler by tracking laser points with PSD. Then on this basis, due to the spatial declination angle in the PSD imaging model, a spatial declination correction model was established to compensate the spatial declination error. Finally, there are errors occurred in the measurement because of the nonlinearity of PSD, and the improved backpropagation (BP) neural network calibration algorithm is adopted to further correct the errors through the Fletcher–Reeves linear search method to obtain the accurate coordinate information of the annunciator. MATLAB is used to verify the convergence of the calibration algorithm error. Simulation results show that 951 epochs need to be trained by using the traditional BP algorithm to make the position error converge to 0.001 mm. But when using the improved BP algorithm, only 44 epochs are needed.

Transient flow behavior in serpentine curved microchannel of inertial microfluidic devices

The serpentine microchannel or bent microchannel has been identified as one of the essential elements for inertial microfluidic devices. Understanding the transient flow of liquid is important in predicting flow time and evaluating device performance. In this paper, based on the unsteady Bernoulli equation, the inertial transient flow model was derived by analyzing the microchannel resistance factor using back propagation (BP) artificial neural network (ANN). In order to train the ANN, an improved BP algorithm was used to make the resistance model more accurate. A centrifugal test platform was employed to measure flow time in the microchannel with different structure sizes. The results show that the transient flow model has a good agreement with the experiment values, the results also show that the method using BP neural network can also be used to solve the problem of nonlinear inertial flow.

A COMPUTER NETWORK INTRUSION DETECTION TECHNOLOGY BASED ON IMPROVED NEURAL NETWORK ALGORITHM

In order to resist the network malicious attack, Back-Propagation (BP) neural network was improved by particle swarm optimization (PSO) algorithm. Then the simulation analysis was carried out in the MATLAB software. The results showed that the improved BP algorithm converged faster and the error was smaller when training the algorithm; compared with BP, PSO-BP had higher accuracy and precision and lower false positive rate, and it also had better detection performance when the size of training samples was small. In summary, PSO-BP can be used for the detection of network intrusion threats.

Decoupling Research of a Novel Three-Dimensional Force Flexible Tactile Sensor Based on an Improved BP Algorithm.

Decoupling research on flexible tactile sensors play a very important role in the intelligent robot skin and tactile-sensing fields. In this paper, an efficient machine learning method based on the improved back-propagation (BP) algorithm is proposed to decouple the mapping relationship between the resistances of force-sensitive conductive pillars and three-dimensional forces for the 6 × 6 novel flexible tactile sensor array. Tactile-sensing principles and numerical experiments are analyzed. The tactile sensor array model accomplishes the decomposition of the force components by its delicate structure, and avoids direct interference among the electrodes of the sensor array. The force components loaded on the tactile sensor are decoupled with a very high precision from the resistance signal by the improved BP algorithm. The decoupling results show that the k-cross validation (k-CV) algorithm is a highly effective method to improve the decoupling precision of force components for the novel tactile sensor. The large dataset with the k-CV method obtains a better decoupling accuracy of the force components than the small dataset. All of the decoupling results are fairly good, and they indicate that the improved BP model with a strong non-linear approaching ability has an efficient and valid performance in decoupling force components for the tactile sensor.

Multilayer Perceptron Based Equalizer with an Improved Back Propagation Algorithm for Nonlinear Channels

Neural network based equalizers can easily compensate channel impairments; such additive noise and inter symbol interference (ISI). The authors present a new approach to improve the training efficiency of the multilayer perceptron (MLP) based equalizer. Their improvement consists on modifying the back propagation (BP) algorithm, by adapting the activation function in addition to the other parameters of the MLP structure. The authors report on experiment results evaluating the performance of the proposed approach namely the back propagation with adaptive activation function (BPAAF) next to the BP algorithm. To further prove its effectiveness, the proposed approach is also compared beside a so known nonlinear equalizer explicitly the multilayer perceptron with decision feedback equalizer MLPDFE. The authors consider various performance measures specifically: signal resorted quality, lower steady state MSE reached and minimum bit error rate (BER) achieved, where nonlinear channel equalization problems are employed.

An improved back propagation algorithm for training neural network-based equaliser for signal restoration in digital communication channels

An improved back propagation algorithm for training neural network-based equaliser for signal restoration in digital communication channels

An improved back propagation algorithm for training neural network-based equaliser for signal restoration in digital communication channels

The back propagation BP algorithm has been very successful in training multilayer perceptron-based equalisers; despite its success BP convergence is still too slow. Within this paper we present a new approach to enhance the training efficiency of the multilayer perceptron-based equaliser MLPE. Our approach consists on modifying the conventional back propagation algorithm, through creating an adaptive nonlinearity in the activation function. Experiment results evaluates the performance of the MLPE trained using the conventional BP and the improved back propagation with adaptive gain IBPAG. Due to the adaptability of the activation function gain the nonlinear capacity and flexibility of the MLP is enhanced significantly. Therefore, the convergence properties of the proposed algorithm are more improved compared to the BP. The proposed algorithm achieves the best performance in the entire simulation experiments.

Dynamic modeling of exergy efficiency of turboprop engine components using hybrid genetic algorithm-artificial neural networks

Genetic algorithm is utilized to design the optimum initial value of parameters and topology of the artificial neural network which is trained by applying the improved backpropagation algorithm using momentum factor so as to minimize the spent time and effort. In this study, a comprehensive dynamic modeling of turboprop engine components plant is accomplished using hybrid GA (genetic algorithm) ANN (artificial neural networks) strategy. The turboprop engine is equipped with main components such as compressor, combustor, gas turbine and power turbine. Newly derived GA-ANN model takes into account five independent engine variables (i.e., torque, power, gas generator speed, engine mass air flow and fuel flow). These dynamic variables are used as inputs of the ANN while exergy efficiencies of the components are considered as the output parameter of the network. The results show that the hybridization with the genetic algorithm has improved the accuracy even further compared to the trial-and-error case, and the estimated values of exergy efficiencies of the components obtained by the derived model provide a close fit with the reference data.

Natural Gas Load Forecasting Based on Improved Back Propagation Neural Network

To suit for the condition that the relative error is more popular than the absolute error, and overcome the shortcoming of the traditional Back propagation neural network, this paper proposed an improved Back propagation algorithm with additional momentum item based on the sum of relative error square. The improved algorithm was applied to the example of the natural gas load forecasting, simulations showed that the improved algorithm has faster training speed than the traditional algorithm, and has higher accuracy as while.

Applied Technology on Artificial Neural Network in Fault Diagnosis System

This paper combined Rumelhart’s adding inertial impulse and dynamically adjusting the learning rate and proposed an improved algorithm to optimize the Back Propagation (BP) networks with applied technology. This improved BP networks is used to determining membership function and applied in fuzzy diagnosing vapor congealing equipment. The application results prove that the improved BP algorithm is effective and the convergence speed is accelerated and is much faster than the classic BP algorithm. The applied technology is very useful in the application course.

Soil infiltration based on bp neural network and grey relational analysis

Soil infiltration is a key link of the natural water cycle process. Studies on soil permeability are conducive for water resources assessment and estimation, runoff regulation and management, soil erosion modeling, nonpoint and point source pollution of farmland, among other aspects. The unequal influence of rainfall duration, rainfall intensity, antecedent soil moisture, vegetation cover, vegetation type, and slope gradient on soil cumulative infiltration was studied under simulated rainfall and different underlying surfaces. We established a six factor-model of soil cumulative infiltration by the improved back propagation (BP)-based artificial neural network algorithm with a momentum term and self-adjusting learning rate. Compared to the multiple nonlinear regression method, the stability and accuracy of the improved BP algorithm was better. Based on the improved BP model, the sensitive index of these six factors on soil cumulative infiltration was investigated. Secondly, the grey relational analysis method was used to individually study grey correlations among these six factors and soil cumulative infiltration. The results of the two methods were very similar. Rainfall duration was the most influential factor, followed by vegetation cover, vegetation type, rainfall intensity and antecedent soil moisture. The effect of slope gradient on soil cumulative infiltration was not significant.

An Improved Gauss-Newtons Method based Back-propagation Algorithm for Fast Convergence

The present work deals with an improved back-propagation algorithm based on Gauss-Newton numerical optimization method for fast convergence. The steepest descent method is used for the back-propagation. The algorithm is tested using various datasets and compared with the steepest descent back-propagation algorithm. In the system, optimization is carried out using multilayer neural network. The efficacy of the proposed method is observed during the training period as it converges quickly for the dataset used in test. The requirement of memory for computing the steps of algorithm is also analyzed.

Target Location Based on Ultrasonic Ranging and Vision Information for Picking Robot

In order to confirm the spatiallocation of the fruit target, a target location method by inosculating ultrasonic ranging and vision was developed for eggplant picking robot. The two dimension image information was achieved by means of color camera. The distance of fruit target was measured by using ultrasonic distance measurer. A three layers BP(Back Propagation) neural network was structured to locate the eggplant. The input variables of the neural network were image center coordinates and distance obtained by ultrasonic distance measurer. The output were space coordinates of picking point. The improved BP algorithm was used to train the parameter of the neural network. The effective parameter was achieved after 100 times of training. Experiments showed that the average rms error of the space coordinates of eggplant was 15.6mm when the measure distance ranges from 150 mm to 600 mm and the average time consumed was 0.27s. The target location method by inosculating ultrasonic ranging and vision information for eggplant picking robot owns good intelligence and wide adaptability and it can meet the demands of the eggplant picking robot.

An Improved Back-Propagation Algorithm for Fuzzy Modeling

Fuzzy modeling is discussed in many literatures and there are numerous algorithms are proposed. Back-propagation algorithm is an efficient algorithm for fuzzy modeling and many papers proposed the usage of such method. But there exists potential risk of dead zone, abrupt inference surface and decreasing sensitivity for normal back-propagation algorithm in fuzzy modeling. This paper analysis the potential problems of normal algorithm and suggest a reformative back-propagation algorithm for fuzzy modeling. A complete algorithm is presented in the paper and some simulate result is discussed