An improved Levenberg-Marquardt algorithm with adaptive learning rate for RBF neural network

An, Ru; Li, Wen Jing; Han, Hong Gui; Qiao, Jun Fei

doi:10.1109/chicc.2016.7553917

Cited by 12 publications

(7 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ANN model consists of an input layer, a hidden layer and an output layer. An adaptive Levenberg-marquardt algorithm is used to train the relationship between the input variables and the measured cooling load of the occupied space [ 48 ]. The root mean squared error (RMSE) and Pearson correlation coefficient (R) are used for the model evaluation.…”

Section: Methodsmentioning

confidence: 99%

A Practical Multi-Sensor Cooling Demand Estimation Approach Based on Visual, Indoor and Outdoor Information Sensing

Wang

Tse

Poon

et al. 2018

Sensors

View full text Add to dashboard Cite

The operating efficiency of heating, ventilation and air conditioning (HVAC) system is critical for building energy performance. Demand-based control is an efficient HVAC operating strategy, which can provide an appropriate level of HVAC services based on the recognition of actual cooling “demand.” The cooling demand primarily relies on the accurate detection of occupancy. The current researches of demand-based HVAC control tend to detect the occupant count using cameras or other sensors, which often impose high computation and costs with limited real-life applications. Instead of detecting the occupant count, this paper proposes to detect the occupancy density. The occupancy density (estimated by image foreground moving pixels) together with the indoor and outdoor information (acquired from existing sensors) are used as inputs to an artificial neural network model for cooling demand estimation. Experiments have been implemented in a university design studio. Results show that, by adding the occupancy density, the cooling demand estimation error is greatly reduced by 67.4% and the R value is improved from 0.75 to 0.96. The proposed approach also features low-cost, computationally efficient, privacy-friendly and easily implementable. It shows good application potentials and can be readily incorporated into existing building management systems for improving energy efficiency.

show abstract

Section: Methodsmentioning

confidence: 99%

A Practical Multi-Sensor Cooling Demand Estimation Approach Based on Visual, Indoor and Outdoor Information Sensing

Wang

Tse

Poon

et al. 2018

Sensors

View full text Add to dashboard Cite

show abstract

“…In order to solve this problem, the improved form of the Gauss-Newton method combined with the LM training algorithm is used to train the network in this paper. The proposed method can accelerate the network training and convergence speed effectively [21,22]. The LM algorithm is a second-order algorithm.…”

Section: Lm Algorithmmentioning

confidence: 99%

Research on the Prediction Method of Centrifugal Pump Performance Based on a Double Hidden Layer BP Neural Network

Han

Nan

et al. 2019

Energies

View full text Add to dashboard Cite

With the aim of improving the shortcomings of the traditional single hidden layer back propagation (BP) neural network structure and learning algorithm, this paper proposes a centrifugal pump performance prediction method based on the combination of the Levenberg–Marquardt (LM) training algorithm and double hidden layer BP neural network. MATLAB was used to establish a double hidden layer BP neural network prediction model to predict the head and efficiency of a centrifugal pump. The average relative error of the head between the experimental and prediction obtained by the double hidden layer BP neural network model was 4.35%, the average relative error of the model prediction efficiency and the experimental efficiency was 2.94%, and the convergence time was 1/42 of that of the single hidden layer. The double hidden layer BP neural network model effectively solves the problems of low learning efficiency and easy convergence into local minima—issues that were common in the traditional single hidden layer BP neural network training. Furthermore, the proposed model realizes hydraulic performance prediction during the design process of a centrifugal pump.

show abstract

“…There is no specific technique or algorithm to find the optimum number of hidden layers or number of neurons in each layer for a certain problem. However, Kamiński (2016) and An et al (2016) presented methods for optimizing the weights, biases and learning rate. To control the temperature of the satellite, passive and active control systems are used.…”

Section: Related Workmentioning

confidence: 99%

Simulation and Prediction for a Satellite Temperature Sensors Based on Artificial Neural Network

Abdelkhalek

Medhat

Ziedan

et al. 2019

J.Aerosp. Technol. Manag.

View full text Add to dashboard Cite

Spacecrafts in space environment are exposed to several kinds of thermal sources such as radiation, albedo and emitted IR from the earth. The thermal control subsystem in spacecraft is used to keep all parts operating within allowable temperature ranges. A failure in one or many temperature sensors could lead to abnormal operation. Consequently, a prediction process must be performed to replace the missing data with estimated values to prevent abnormal behavior. The goal of the proposed model is to predict the failed or missing sensor readings based on artificial neural networks (ANN). It has been applied to EgyptSat-1 satellite. A backpropagation algorithm called Levenberg-Marquardt is used to train the neural networks (NN). The proposed model has been tested by one and two hidden layers. Practical metrics such as mean square error, mean absolute error and the maximum error are used to measure the performance of the proposed network. The results showed that the proposed model predicted the values of one failed sensor with adequate accuracy. It has been employed for predicting the values of two failed sensors with an acceptable mean square and mean absolute errors; whereas the maximum error for the two failed sensors exceeded the acceptable limits.

show abstract

An improved Levenberg-Marquardt algorithm with adaptive learning rate for RBF neural network

Cited by 12 publications

References 29 publications

A Practical Multi-Sensor Cooling Demand Estimation Approach Based on Visual, Indoor and Outdoor Information Sensing

A Practical Multi-Sensor Cooling Demand Estimation Approach Based on Visual, Indoor and Outdoor Information Sensing

Research on the Prediction Method of Centrifugal Pump Performance Based on a Double Hidden Layer BP Neural Network

Simulation and Prediction for a Satellite Temperature Sensors Based on Artificial Neural Network

Contact Info

Product

Resources

About