Performance Degradation Prediction Using LSTM with Optimized Parameters.

Predicting the degradation of mechanical components, such as rolling bearings is critical to the proper monitoring of the condition of mechanical equipment. A new method, based on a long short-term memory network (LSTM) algorithm, has been developed to improve the accuracy of degradation prediction. The model parameters are optimized via improved particle swarm optimization (IPSO). Regarding how this applies to the rolling bearings, firstly, multi-dimension feature parameters are extracted from the bearing's vibration signals and fused into responsive features by using the kernel joint approximate diagonalization of eigen-matrices (KJADE) method. Then, the between-class and within-class scatter (SS) are calculated to develop performance degradation indicators. Since network model parameters influence the predictive accuracy of the LSTM model, an IPSO algorithm is used to obtain the optimal prediction model via the LSTM model parameters' optimization. Finally, the LSTM model, with said optimal parameters, was used to predict the degradation trend of the bearing's performance. The experiment's results show that the proposed method can effectively identify the trends of degradation and performance. Moreover, the predictive accuracy of this proposed method is greater than that of the extreme learning machine (ELM) and support vector regression (SVR), which are the algorithms conventionally used in degradation modeling.

Optimization of virotherapy for cancer.

Several viruses preferentially infect and replicate in cancer cells by usurping pathways that are defective in the tumor cell population. Such viruses have a potential as oncolytic agents. The aim of tumor virotherapy is that after injection of the replicating virus, it propagates in the tumor cell population with amplification. As a result, the oncolytic virus spreads to eradicate the tumor. The outcome of tumor virotherapy is determined by population dynamics and different from standard cancer therapy. Several models have been developed that provided considerable insights on the potential therapeutic scenarios. However, virotherapy is potentially risky since large amounts of a replicating virus are injected in the host with a risk of adverse effects. Therefore, the optimal dose, number of doses, and timing are expected to play an important role on the outcome both for the tumor and the host. In the current work, we combine a model of the dynamics of tumor virotherapy that was validated with experimental data with optimization theory to illustrate how we can improve the outcome of tumor therapy. In this first report, we demonstrate that (i) in most circumstances, anything more than two administrations of a vector is not helpful, (ii) correctly timed delivery of the virus provides superior results compared to regularly scheduled therapy or continuous infusion, (iii) a second dose of virus that is not properly timed leads to a worse outcome compared to a single dose of virus, and (iv) it is less costly to treat larger tumors.

Binder Free Hierarchical Mesoporous Carbon Foam for High Performance Lithium Ion Battery.

A hierarchical mesoporous carbon foam (ECF) with an interconnected micro-/mesoporous architecture was prepared and used as a binder-free, low-cost, high-performance anode for lithium ion batteries. Due to its high specific surface area (980.6 m2/g), high porosity (99.6%), light weight (5 mg/cm3) and narrow pore size distribution (~2 to 5 nm), the ECF anode exhibited a high reversible specific capacity of 455 mAh/g. Experimental results also demonstrated that the anode thickness significantly influence the specific capacity of the battery. Meanwhile, the ECF anode retained a high rate performance and an excellent cycling performance approaching 100% of its initial capacity over 300 cycles at 0.1 A/g. In addition, no binders, carbon additives or current collectors are added to the ECF based cells that will increase the total weight of devices. The high electrochemical performance was mainly attributed to the combined favorable hierarchical structures which can facilitate the Li+ accessibility and also enable the fast diffusion of electron into the electrode during the charge and discharge process. The synthesis process used to make this elastic carbon foam is readily scalable to industrial applications in energy storage devices such as li-ion battery and supercapacitor.