Mechanical seal is a kind of shaft sealing equipment. Face wear is one of the main causes of mechanical seal failure. Mechanical seal condition is also related to the reduction of energy consumption and carbon emission. Therefore, we need to detect the centrifugal pump seal condition. At present, vibration signal is a common method for fault monitoring and diagnosis of centrifugal pump. In this paper, the vibration signal under the condition of damaged centrifugal pump seal is measured by studying the characteristics of vibration signal after the end face damage of centrifugal pump. Statistical indicators such as RMS and kurtosis were taken to analyze the average energy and shock wave energy of vibration signal. The time‐frequency characteristics of vibration signal are analyzed by frequency spectrum. The results show that there are a large extent variation of vibration amplitude in the direction of base and axis and a weak variation of vibration amplitude in the direction of radial and vertical. With the increasing of flow rate, the RMS of vibration signal falls at first, then keeps steady, and mounts at last when the flow rate is over the design flow rate. It can be shown from the time‐frequency spectrum that there is a shock wave and pause signals caused by the shock wave, which are reflected by the higher frequency band components of the vibration signal that can provide a reference to the diagnosis of the occurrence of damaged mechanical seal. From the analysis, the energy of vibration signal is related to the running condition, we can find that the occurrence of mechanical seal wear makes the centrifugal pump to produce high‐frequency vibration signal, and the axial vibration is the strongest and the instability in the fluid makes the vibration signal produce high amplitude characteristics. Analyzing the vibration signal characteristics of centrifugal pumps with damaged mechanical seal is of great significance to find the mechanical seal failure of the centrifugal pumps and adjust the operating parameters.

With extensively using of centrifugal pumps, noise generation in these pumps is increasingly receiving research attention in recent years. The noise sources in centrifugal pumps are mainly composed of mechanical noise and flow-induced noise. And the study of flow-induced noise has become a hotspot and important domain in the field. The flow-induced noise closely related to the inner pressure pulses and vibration of volute in pumps, therefore, it is necessary to research the interaction and mechanism among them. To investigate the relationships, a test system is designed which includes a test loop and a measurement system. The hydrophones and pressure sensors are installed on the outlet of the pump and vibration acceleration sensors are disposed on the pump body. Via these instruments, the signals of noise, pressure pulses and vibration are collected and analyzed. The results show that the level of flow-induced noise becomes smaller as the flow increment during low flow rate operations, and it is steadily close to the design point, then it increases with the growing of flow rate in high flow rate conditions. Furthermore, there are some similar peak points in the power spectrum charts of noise, pressure pulses and vibration. The broadband noise at low flow rate is mostly focused on the region of 0–40 times shaft frequency, which is mostly made by rotating stall and vortex; while the noise at high flow rate conditions is focused on the region of 60–100 times shaft frequency, which may be mostly made by cavitations. The proposed research is of practical and academic significance to the study of noise reduction for centrifugal pumps.

To effectively predict the faults of centrifugal pumps, the idea of machine learning k‐nearest neighbor algorithm (KNN) was introduced into the traditional Mahalanobis distance fault discrimination, and an improved centrifugal pump fault prediction model of KNN based on the Mahalanobis distance is proposed. In this method, the Mahalanobis distance is used to replace the distance function in the conventional KNN algorithm. Grid search and cross‐validation are used to determine the optimal K value of the prediction model. A centrifugal pump test rig was established to solve three common faults of centrifugal pumps: cavitation, impeller damage, and machine seal damage, and the method was verified. The results show that this method can effectively distinguish the specific fault types of centrifugal pumps based on vibration signals, and the fault prediction accuracy of the off‐balance condition is up to 82%. This study provides a novel idea and method for centrifugal pump fault prediction and diagnosis and avoids the interaction between parameters when monitoring multiple parameters.

Sensorless monitoring technology based on motor current signature analysis is a nonintrusive and economical technique to monitor motor-driven equipment. Sensorless monitoring technology can be applied to a centrifugal pump system. This technology is also based on the motor current signature of centrifugal pump load; however, systematic research regarding motor current signature in overall normal operation points which is the applied basic for sensorless monitoring technology has been rarely performed. As such, we partially examined the motor current signature of a centrifugal pump load by experimental observation, theoretical analysis, and numerical simulation. Results show that stator current is a sinusoidal alternating current that strictly follows sine law associated with the cycle of the fundamental frequency of supply power. The trend of the root mean square and peak–peak of current is the same as flow–shaft power characteristics; hence, this trend could be used as indicator of the pump operational point monitor. The frequency characteristics of a centrifugal pump, such as blade passing frequency, rotation frequency, and broadband noise, could be reflected as sidebands around the fundamental frequency. The stator current spectrum is composed of fundamental frequency component, harmonics component, and noise. The fundamental frequency component is directly related to the pump load in which changes associated with the law of fundamental frequency component are relatively similar to flow–shaft power characteristics. Harmonics component and noise are caused by load fluctuation in which the amount of energy of these two components exhibits a lower value at the preferred operation point. By contrast, the amount of energy likely increases when pump operation is at an unstable operation point. These results further indicate that motor current signature analysis is a feasible and cost-effective method to monitor centrifugal pump operation status. Therefore, motor current signature analysis can be applied to monitor-related flow phenomena.

Cavitation is a major undesirable phenomenon for centrifugal pump because it can cause hydraulic performance deterioration, pump damage by pitting and material erosion, and structural vibration and noise. Cavitation can appear within the entire range of the operating conditions; therefore, it must be prevented by all means. Sensorless monitoring technology based on motor current signature analysis is non-intrusive and economic for monitoring motor-driven equipment. Thus, this technology is suitable for centrifugal pump systems. The motor current signature for centrifugal pump load at the cavitation condition is the basis of this technology. However, systematic research is lacking on sensorless monitoring technology based on motor current signature. As a result, the tentative exploration for motor current signature at cavitation load was conducted in this study. The results show that the stator current is still a sinusoidal alternating current strictly to the law of sine. Moreover, the root mean square of the current fluctuates because of different flow regimes in the cavitation progress and decreases because vapor density is smaller than water density when cavitation is fully formed. For the stator current spectrum, the noise level, noise distribution, rotation speed, and vane pass frequency components show features in the cavitation process. These indicator indexes change according to the stage of cavitation development. Thus, the motor current signature analysis is found to be a feasible and cost-effective method for the stages of cavitation condition.

Centrifugal pumps are essential in many industrial processes. An accurate operation diagnosis of centrifugal pumps is crucial to ensure their reliable operation and extend their useful life. In real industry applications, many centrifugal pumps lack flowmeters and accurate pressure sensors, and therefore, it is not possible to determine whether the pump is operating near its best efficiency point (BEP). This paper investigates the detection of off-design operation and cavitation for centrifugal pumps with accelerometers and current sensors. To this end, a centrifugal pump was tested under off-design conditions and various levels of cavitation. A three-axis accelerometer and three Hall-effect current sensors were used to collect vibration and stator current signals simultaneously under each state. Both kinds of signals were evaluated for their effectiveness in operation diagnosis. Signal processing methods, including wavelet threshold function, variational mode decomposition (VMD), Park vector modulus transformation, and a marginal spectrum were introduced for feature extraction. Seven families of machine learning-based classification algorithms were evaluated for their performance when used for off-design and cavitation identification. The obtained results, using both types of signals, prove the effectiveness of both approaches and the advantages of combining them in achieving the most reliable operation diagnosis results for centrifugal pumps.