Liu Datong

Online adaptive status prediction strategy for data-driven fault prognostics of complex systems

Accurate fault prediction or remaining useful life (RUL) estimation can obviously reduce cost of maintenance and decrease the probability of accidents so as to improve the performance of the system test and maintenance. At the same time, it is difficult to apply the model-based methods for fault prediction in most applications for complex systems. Due to continuously improving of automation, increasing of sampling frequency and development of computing technology and memory capacity, it gradually promotes data-driven technology into practical methods. Therefore, data-driven fault prognostics based on the sensor or historical test data has become the primary prediction means of complex systems, such as Artificial Neural Networks (ANN), Support Vector Regression (SVR) and other computational intelligence methods. In the other hands, most of traditional forecasting methods are always off-line that are not suitable for on-line prediction and real-time processing. Furthermore, for some on-line prediction methods such as Online Support Vector Regression (Online SVR), there is conflicts and trade-offs between prediction efficiency and accuracy. In real prognostics and health management (PHM) systems, such as the operating status monitoring and forecasting of complex systems like airplane and aircraft, it requires that the algorithms are flexible and adaptive in realization of balance between prediction efficiency and accuracy to meet different complicated requirements. To solve the problem above, an on-line adaptive data-driven fault prognosis and prediction strategy is presented in this paper. Considering the complex characteristics of trends and neighborhood of time series data, the multi-scale reconstruction strategy is applied to effectively reduce the size of on-line data and preserve rich history knowledge of samples. Therefore, the prediction efficiency could be improved and faster forecasting can be achieved with adaptive multi-scale sub-models. To evaluate the proposed prediction strategy, we have executed experiments with Tennessee Eastman (TE) process data. Experimental results with TE process fault data prove its effectiveness. The experiments and tests confirm the algorithms can be effectively applied to the on-line status monitoring and prediction with excellent performance in both efficiency and precision. New on-line fault status prediction strategy shows better prospect in real-time and on-line application for complex system. It can be applied in industrial fields for system maintenance and prognostics and health management.

Fault prediction based on time series with online combined kernel SVR methods

In order to reduce the cost and decrease the probability of accidents, accurate fault prediction is a goal pursued by researchers working at system test and maintenance. Most of traditional fault forecasting methods are not suitable for online prediction and real-time processing. To solve this problem, an online data-driven fault prognosis and prediction method is presented in this paper. The operating states are forecasted with on-line time series prediction model based on the online combined kernel functions Support Vector Regression (SVR). Compared with batch SVR prediction models, online SVR has a good real-time processing performance. However, it is hard for a single kernel SVR to obtain accurate result for the complicated nonlinear and non-stationary time series. Therefore, a combined online SVR with different kernels containing global and local kernels is developed for fault prediction. For general fault modes, the fault trend feature can be extracted by global kernel. On the other hand, local kernel can reflect and revise the local changes of data characteristics in neighborhood. It has realized better result than the method of the single SVR. Experimental results for Tennessee Eastman process fault data prove its effectiveness.

Data-driven framework for lithium-ion battery remaining useful life estimation based on improved nonlinear degradation factor

This paper proposes an improved nonlinear degradation factor based on the current percentage of life-cycle length (CPLL) which contains the battery capacity degradation characteristics information of different periods. This method is improved based on related nonlinear degradation Autoregressive (AR) data-driven prognostics model considering an improved scale nonlinear degradation factor. Then a combination is implemented between the proposed factor and data-driven AR model named nonlinear scale degradation parameter based AR (NSDP-AR) model for better nonlinear prediction ability. Extended Kalman Filter (EKF) is used to obtain the specific factor for certain kind of battery. In order to promote the modified model, a remaining useful life (RUL) prognostic framework using Grey Correlation Analysis (GCA) will be established. The experimental results with the battery data sets from NASA PCoE and CALCE show that the proposed NSDP-AR model and the corresponding prognostic framework can achieve satisfied RUL prediction performance.

Telemetry-data based anomaly detection method for flywheel of in-orbit satellite

Due to that some parameters are unable to be measured in orbit, it is difficult to establish a precise physical model for in-orbit satellite flywheel. Aiming at this issue, a telemetry-data-based model for anomaly detection is proposed. The model considers only the main parameters that are closely related to the operating condition of the flywheel. Then, the prediction model is trained and tested based on normal telemetry data. The test result indicates that the predictors are consistent with the actual updating value. Moreover, an in-orbit flywheel fault detection method is proposed based on the predicted model. Finally, the method is verified with an in-orbit flywheel failure case, the experimental results indicate that the proposed method can timely and accurately identify the failure.

Fault diagnosis for discrete monitoring data based on fusion algorithm

Fault diagnosis has a significant role in enhancing the safety, reliability, and availability of complex systems. However, the problem of enormous condition monitoring data and multiple failure modes makes the diagnostics great challenge. The imbalance between normal and fault monitoring data will increase the false alarm rate and the false negative rate. On the other hand, discrete monitoring data such as events are frequent and critical to fault diagnosis of complex systems. In this work, we propose a fusion fault diagnostic method which combines Naïve Bayes with AdaBoost ensemble algorithm. This integrated method is appropriate for discrete data and improves the adaptability for imbalanced condition monitoring data. Experimental results based on PHM 2013 dataset show that fault diagnosis performance using the fusion method can be ameliorated.

Optimized FPGA-based DDR2 SDRAM controller

With the development of embedded systems, more and more applications require large and high speed memory. The FPGA-based solution also faces the same demand. Design and realization of an external storage with large capacity and high throughput in the FPGA-based embedded system is becoming a challenge. To satisfy the practical requirement, a DDR2 controller design is proposed, which efficiently and selectively integrates with the Altera DDR2 SDRAM High Performance Controller (HPC) module. Finally, the optimized DDR2 SDRAM controller based on Altera HPC is realized, and the goal that data accesses for DDR2 SDRAM with the ability of increase channel and relatively huge burst size is achieved. The optimized DDR2 controller has been implemented and verified in the Altera EP2SGX90E FPGA, and revealed a significant improvement in the performance compared with the individual HPC module. The experimental results show that this optimized DDR2 SDRAM controller demonstrates the properties of multichannel and high bandwidth memory access.

Design of a PCI Bus High Precision and Wide Range Programmable Current Generator

Corresponding to the common dilemma of programmable current generators narrow output range and low precision, this paper is intended to present one design of a PCI bus high precision and wide programmable current generator. A 16-bit DAC is put into use to improve the precision of output voltage and V-I transition is easily realized by means of the operational amplifiers which has features such as low noise, temperature drift, distortion and high driving ability. By the precise sampling resistors and 24-bit ADC, the design samples the output current. And then based on the results, discrete increment PID close loop control is adopted in the drivers to realize self calibration of the output. Finally, validated by HP 34401A millimeter, the results shows the output ranges from -50mA to +50mA when the output error is smaller than plusmn3muA, which proves that the design improves output range and precision.