Fault Diagnosis and Condition Monitoring of Reciprocating Compressors

Reciprocating compressors are pivotal in industries ranging from oil and gas to chemical processing, where reliable pressurisation of gases underpins safety and productivity. Their mechanical complexity—including pistons, valves and bearings—inevitably gives rise to faults such as valve fluttering, piston ring wear and bearing defects. Condition monitoring aims to detect deviations from normal operation through continuous measurement of parameters like vibration, pressure and temperature. Fault diagnosis then interprets these measurements to identify the underlying malfunction. The principal challenges lie in the nonlinear, non-stationary nature of compressor signals, the influence of operating conditions on fault signatures and the need for rapid, automated analysis. Recent advances blend classical signal-processing techniques—wavelet transforms, adaptive decomposition and statistical feature extraction—with data-driven methods such as machine learning and deep neural networks. These approaches have achieved high diagnostic accuracy by fusing multisource data, exploiting novel image-based thermal patterns and incorporating explainable anomaly detection. Together, they support predictive maintenance strategies that reduce unscheduled downtime and extend equipment life spans.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Fault Diagnosis and Condition Monitoring of Reciprocating Compressors publication trend

The graph below shows the total number of articles in fault diagnosis and condition monitoring of reciprocating compressors across all publications each year (not limited to Nature Index journals).

Technical terms

Reciprocating compressor: A positive-displacement machine that uses back-and-forth piston motion to compress gas.

Condition monitoring: The continuous or periodic measurement of operational parameters to assess equipment health.

Fault diagnosis: The process of interpreting condition-monitoring data to identify specific mechanical or functional anomalies.

Vibration signal: Oscillatory data collected from sensors that reflect mechanical movements and potential faults.

Feature fusion: The integration of multiple data representations to form a comprehensive set of inputs for diagnostic models.

Convolutional neural network (CNN): A deep-learning architecture that automatically learns spatial or temporal patterns from input data.

Stack denoising autoencoder (SDAE): A neural network trained to reconstruct input data from a corrupted version, enabling robust feature extraction.

References

1. Prediction of air compressor faults with feature fusion and machine learning. Knowledge-Based Systems (2024).
2. An Automatic Fault Diagnosis Method for the Reciprocating Compressor Based on HMT and ANN. Applied Sciences (2022).
3. An Intelligent Fault Diagnosis Method for Reciprocating Compressors Based on LMD and SDAE. Sensors (2019).