Hybridization of harmonic search algorithm in training radial basis function with dynamic decay adjustment for condition monitoring

Abstract

In recent decades, hybridization of superior attributes of few algorithms was proposed to aid in covering more areas of complex application as well as improves performance. This paper presents an intelligent system integrating a Radial Basis Function Network with Dynamic Decay Adjustment (RBFN-DDA) with a Harmony Search (HS) to perform condition monitoring in industrial processes. An effective condition monitoring can help reduce unexpected breakdown incidents and facilitate in maintenance. RBDN-DDA performs incremental learning wherein its structure expands by adding new hidden units to include new information. As such, its training can reach stability in a shorter time compared to the gradient-descent based methods. By integrating with the HS algorithm, the proposed metaheuristic neural network (RBFN-DDA-HS) can optimize the RBFN-DDA parameters and improve classification performances from the original RBFN-DDA by 2.2% up to 22.5% in two benchmarks and a real-world condition-monitoring case studies. The results also show that the proposed RBFN-DDA-HS is compatible, if not better than, the classification performances of other state-of-art machine learning methods.

Review of fault detection techniques for predictive maintenance

“…, 2021). Condition monitoring using ANN models assures better accuracy in production systems (Chong et al. , 2021).…”

“…In the manufacturing industry, many sensor devices are interconnected to collect operational data from machines on a continuous basis and feed it to backend computers for control and predictive analysis (Gopalakrishnan and Kumaran, 2022;Masero et al, 2018). These techniques can be used to monitor different parts of a manufacturing process (He et al, 2017), including belt drives (Pollak et al, 2021), bearings (Pichler et al, 2020;Lalik and Wa ˛torek, 2021), fleet (Ioanna et al, 2021), boiler feed pumps (Moleda et al, 2020), bi-directional JQME 29,2 control valve (Khadim et al, 2021), rotating machinery (Torres-Contreras et al, 2021;Lis et al, 2021), conveyor motors (Kiangala and Wang, 2020), injection moulding machines (Rousopoulou et al, 2020) and steel plate systems (Chong et al, 2021). In addition to all these, failures can happen in key components such as barrier machines in railroad crossings (Grzechca et al, 2021a) and heavy Earth moving machinery.…”

“…Condition monitoring using ANN models assures better accuracy in production systems (Chong et al, 2021). With the aid of neural networks, it is possible to anticipate motor oil conditions (Rodrigues et al, 2020;Ullah et al, 2017) and photovoltaic system maintenance (Benedetti et al, 2018) in the electrical industry.…”

“…, 2021), conveyor motors (Kiangala and Wang, 2020), injection moulding machines (Rousopoulou et al. , 2020) and steel plate systems (Chong et al. , 2021).…”