The unsteady and discontinuous liquid flow in the microchannel affects the efficiency of sample mixing, molecular detection, target acquisition, and biochemical reaction. In this work, an active method of reducing the flow pulsation in the microchannel of a pneumatic microfluidic chip is proposed by using an on-chip membrane microvalve as a valve chamber damping hole or a valve chamber accumulator. The structure, working principle, and multi-physical model of the reducing element of reducing the flow pulsation in a microchannel are presented. When the flow pulsation in the microchannel is sinusoidal, square wave, or pulse, the simulation effect of flow pulsation reduction is given when the membrane valve has different permutations and combinations. The experimental results show that the inlet flow of the reducing element is a square wave pulsation with an amplitude of 0.1 mL/s and a period of 2 s, the outlet flow of the reducing element is assisted by 0.017 and the fluctuation frequency is accompanied by a decrease. The test data and simulation results verify the rationality of the flow reduction element in the membrane valve microchannel, the correctness of the theoretical model, and the practicability of the specific application, which provides a higher precision automatic control technology for the microfluidic chip with high integration and complex reaction function.

In this paper, a double-mode self-calibration tension system is proposed, which adopts the conversion of hydraulic meter tension and the monitoring of standard force sensors. According to the material characteristics of the jack and the viscosity and temperature characteristics of the hydraulic oil, the differential model of heat conduction in the hydraulic cylinder and the mathematical model of oil film friction heat generation are established, and the internal thermodynamic characteristics of the jack are theoretically analyzed, which provides theoretical support for the temperature compensation of the hydraulic oil pressure gauge of the jack. A simulation analysis was conducted on the thermodynamic characteristics of the hydraulic jack, and the distribution patterns of the temperature field, thermal stress field, and thermal strain field inside the hydraulic cylinder during normal operation were determined by measuring the temperature changes in five different parts of the jack at different times (t = 200 s, 2600 s, 5000 s, 7400 s, and 10,000 s). For the issue of heat generation due to oil film friction in the hydraulic jack, a simulation calculation model is developed by integrating Computational Fluid Dynamics (CFD) techniques with dynamic grid and slip grid methods. By simulating and analyzing frictional heating under conditions where the inlet pressures are 0.1 MPa, 0.3 MPa, 0.5 MPa, 0.7 MPa, and 0.9 MPa, respectively, we can obtain the temperature distribution on the jack, determine the frictional resistance, and subsequently conduct a theoretical analysis of the simulation results. Using the high-precision standard force sensor after data processing and the hydraulic oil gauge after temperature compensation, the online self-calibration of the tensioning system is carried out, and the regression equation of the tensioning system under different oil temperatures is obtained. The double-mode self-calibration tensioning system with temperature compensation is used to verify the compensation accuracy of the proposed double-mode self-calibration tensioning system.

Reliable fault diagnosis of rotating machines under extreme conditions—strong speed, load variation, intense noise, and severe class imbalance—remains a critical industrial challenge. We develop an ultra-light yet robust framework to accurately detect weak bearing, and gear faults when less than 5% labels, 10 dB noise, 100:1 imbalance and plus or minus 20% operating-point drift coexist. Methods: The proposed Adaptive Feature Module–Conditional Dynamic GRU Auto-Encoder (AFM-CDGAE) first compresses 512 d spectra into 32/48 d “feature modules” via K-means while retaining 98.4% fault energy. A workload-adaptive multi-scale convolution with spatial attention and CPU-aware λ-scaling suppresses noise and adapts to edge–device load. A GRU-based auto-encoder, enhanced by self-attention, is trained with balanced-subset sampling and minority-F1-weighted voting to counter extreme imbalance. On Paderborn (5-class) and CWRU (7-class) benchmarks, the 0.87 M-parameter model achieves 99.12% and 98.83% Macro-F1, surpassing five recent baselines by 3.1% under normal and 5.4% under the above extreme conditions, with only 1.5 to 1.8% F1 drop versus 6.7% for baselines. AFM-CDGAE delivers state-of-the-art accuracy, minimal footprint and strong robustness, enabling real-time deployment at the edge.