Hidden Markov model-based ensemble methods for offline handwritten text line recognition

This paper investigates various ensemble methods for offline handwritten text line recognition. To obtain ensembles of recognisers, we implement bagging, random feature subspace, and language model variation methods. For the combination, the word sequences returned by the individual ensemble members are first aligned. Then a confidence-based voting strategy determines the final word sequence. A number of confidence measures based on normalised likelihoods and alternative candidates are evaluated. Experiments show that the proposed ensemble methods can improve the recognition accuracy over an optimised single reference recogniser.     

Facial synthesis of dandelion-like g-C3N4/Ag with high performance of photocatalytic hydrogen production

Nowadays, energy shortage is one of the major problems in the world. Photocatalytic hydrogen production is a new type of energy technology with good application prospect. As a new type of photocatalytic semiconductor material, g-C₃N₄ has attracted much attention as a photocatalyst. By ultrasonic treatment of a mixed solution of g-C₃N₄ and bovine serum albumin, followed by adding a certain amount of silver nitrate solution and then directly hydrothermal treatment, a special dandelion-like g-C₃N₄/Ag (D-g-C₃N₄/Ag) was prepared. The scanning electronic microscopy, transmission electronic microscopy, X-ray diffraction, Fourier transform infrared, fluorescence and physicochemical adsorption methods were used to characterize the morphology and structure of D-g-C₃N₄/Ag. In addition, the photocatalytic H₂ production of D-g-C₃N₄/Ag with different Ag loadings or in different sacrificial agents and different pH conditions were investigated. The results indicated that when triethanolamine was used as sacrificial agent, photocatalytic hydrogen efficiency was the best, and the rate of photocatalytic hydrogen production reached 862 μmol g⁻¹ h^{− 1} as the Ag loading was 4%.     

Comparative analysis on temperature characteristics of hydrogen-powered and traditional fossil-fueled vehicle fires in the tunnel under longitudinal ventilations

Vehicle fires in the tunnel are a great threat to the safe operation of the tunnel. Due to the rapid development of the hydrogen economy, the fire due to the hydrogen leakage could not be avoided and may bring great damage to the passengers and infrastructure. Due to the large difference between pool fires of traditional fossil-fueled and jet fires of hydrogen-powered vehicles, it is in doubt whether the existing longitudinal ventilation design could still be effective for the safety issue of hydrogen powered vehicles. To solve this problem, it is necessary to compare temperature characteristics of hydrogen-powered and traditional vehicle fires with and without longitudinal ventilations. In present work, we conducted a numerical investigation to discuss the different temperature distributions of traditional and hydrogen-fueled vehicle fires. Results indicate that the high temperature zone of the pool fire only exists above the ceiling of the vehicle. For hydrogen-powered vehicle fire, the high-speed hydrogen jet with the strong inertial force could push the hot smoke flows back to the ground. The ceiling temperature of hydrogen-powered vehicle fire is larger since hydrogen-powered vehicle has a larger heat release rate and the fire hazard of jet fires bring more danger compared with the pool fire. Although the temperature stratification is also obvious for the hydrogen-powered vehicle fire, the air temperature in the lower region could be heated and still high enough to bring a great damage to the passengers’ lives. This is quite different with the traditional pool fire. In addition, the critical ventilation velocity is also discussed. The theoretical equation could well predicted the critical ventilation velocity of traditional vehicle fires. For hydrogen-powered vehicle fires, the critical ventilation velocity could reach up to 6 m/s. The theoretical equation could not well predict the critical ventilation velocity of hydrogen-powered vehicle fires due to exist of hydrogen jet fires.     

Flocculation and mechanism of self-flocculating lipid producer microalga Scenedesmus quadricauda for biomass harvesting

In this study, the extracellular polymeric substance of Scenedesmus quadricauda called EPS-S.q and its bioflocculatoin mechanism were investigated. Results showed EPS-S.q was successfully used as bioflocculating agent for S. quadricauda biomass harvesting and flocculation efficiency of up to 86.7% to S. quadricauda cells could be achieved in presence of Zn²⁺. EPS-S.q was the flocculating agent for self-flocculating microalga S. quadricauda and bioflocculation mechanism was polymer bridging. The sugar and protein mass fraction of dry EPS-S.q was sugar 56.7% and protein 41%. The infrared spectrum further indicated the presence of hydroxyl, carboxyl and amino groups. Moreover, pH decrease induced the flocculation of S. quadricauda and 78.4% of flocculation efficiency was the highest at pH 3. In addition, chemical flocculant FeCl₃ was efficiently used for S. quadricauda harvesting and up to 96.8% of flocculation efficiency could be achieved for S. quadricauda culture with biomass concentration 0.21–0.39 g L⁻¹ at pH 7.     

邛崃1井氮气钻井事故分析(Ⅰ)——构成事故的重要事件

位于四川盆地西部的邛崃1井在侏罗系沙溪庙组地层实施氮气钻井时发生了恶性井喷爆燃事故,这是发生在我国的第一起氮气钻井事故,对尚处于起步阶段的气体钻井技术的发展产生了很大的负面影响,并已经导致了气体钻井应用的大幅度减少。因此,对该次事故进行系统分析并形成正确的认识将是恢复气体钻井技术良性发展的紧迫需要。为此,首先证明了地层产气量过大并非该井事故的根本原因,然后推测了构成事故的重要事件;首次发现气体钻井钻穿裂缝性气层时由于岩石突然爆碎而引发的岩石和天然气猛烈喷出的现象,并将这种现象命名为气体钻井中的"气炮式岩爆井喷",参考矿山行业术语简称"岩爆"或"岩炮"。研究结果表明:①在裂缝性致密砂岩气层中实施气体钻井时的"岩爆"现象是第一位的重要事件或根源事件;②岩爆的大量碎屑和高压气体喷入井内,并在井内反复产生堵塞和卡钻,是岩爆的诱发事件;③然而事故最直接的重要事件却是岩爆碎屑在排砂管线内的堵塞和由此产生的井口高压;④井口高压使得排砂管线爆裂并引发天然气喷出和爆燃,这是事故的直接表象。     

Loss characteristics and flow rectification property of diffuser valves for micropump applications

The primary motivation for this work is to provide data on the loss characteristics and flow rectification performance of flat-walled microdiffuser valves for Reynolds numbers between 100 and 2000 which is considered deficient in literature. The diffusers are designed with a fixed slenderness of 15 and an aspect ratio of 0.92 and with diverging angles ranging from 4° to 120°. The total pressure loss coefficient of the diffuser flow is determined experimentally. Components of the loss are studied in detail by numerical simulations. Results from the computations are in satisfactory agreement with the measurements. Simulations are also carried out for nozzle flows. Based on the experimental and the computational results, efficiencies of the diffuser valves are calculated. It is shown that, for laminar flows, global flow separation plays a significant role in reducing the loss of the diffuser. Consequently, the diffuser angle corresponding to the optimum diffuser efficiency varies from 40° at Re = 100 to 20° for Re ?500.     

Visualisation of pyrite leaching by selected thermophilic archaea: Nature of microorganism–ore interactions during bioleaching

In this study, pyrite (FeS₂) was leached by Acidianus brierleyi, Metallosphaera sedula and Sulfolobus metallicus during a 60 day experimental period. Leaching occurred over a redox potential range of 800 to 860 mV (S.H.E.) and in the presence of increasing Fe³⁺ levels. A modified ferrozine assay was developed to detect the increase of iron in solution as bioleaching of the ore progressed. For the first time, the interactions of these extreme thermophiles with the metal sulfide ore particles were extensively documented using SEM and TEM. As the pyrite degraded, there appeared to be a progression of deposited structures forming, ranging from sub-micron precipitates and disc-shaped structures on the ore's surface, which ultimately were similar for all leaching cultures. Furthermore, the residues resulting from the leaching of pyrite by M. sedula, the most active thermophile, were characterised using SEM/EDX, and appeared to be dominated by iron sulfate precipitates. The nature of the deposits formed, together with our other results, indicate that A. brierleyi, M. sedula and S. metallicus acted through the ‘contact’ and ‘non-contact’ sub-mechanisms of the indirect bioleaching mechanism for the dissolution of pyrite. The role of the bioleaching microorganisms is thus to maintain sufficient levels of Fe³⁺ and acid during pyrite leaching, for maximal mineral dissolution.     

Prediction model for methanation reaction conditions based on a state transition simulated annealing algorithm optimized extreme learning machine

Methanation is the core process of synthetic natural gas, the performance of the entire reaction system depends on precise values of the reaction condition parameters. Accurate predictions of the CO conversion rate of the methanation reaction can eliminate time-consuming and complex steps in experiments and speed up the discovery of the best reaction conditions. However, the methanation reaction is an uncertain, highly complex, and highly nonlinear process. Thus, this paper proposes a machine learning prediction model for the methanation reaction to facilitate the subsequent search for optimal reaction conditions. The reaction temperature, pressure, hydrogen–carbon ratio, water vapor content, CO₂ content, and space velocity were selected as the condition variables. The CO conversion rate was the optimization objective. An extreme learning machine (ELM) was selected as a prediction model. Because the input weights and bias matrices of the ELM are randomly generated, an ELM based on a state transition simulated annealing (STASA-ELM) algorithm is proposed. The STASA algorithm was used to optimize the ELM to improve the accuracy and stability of the model. Five additional sets of experimental data were designed for the experiment, and the error between the experimental and predicted values was small. Thus, the STASA-ELM algorithm can accurately predict the conversion of CO for different values of reaction conditions.     

A comprehensive mathematical model of a serial composite process for biomass and coal Co-gasification

A comprehensive mathematical model to simulate a serial composite process for biomass and coal co-gasification has been built. The process is divided into combustion stage and gasification stage in the same gasifier, it is a new process for the co-gasification of biomass and coal. The model is based on reaction kinetic, hydrodynamics, mass and energy balances, it is a one-dimensional, K-L three-phase, unsteady state model. The model is divided into two sub-models, one is the combustion sub-model, the other is the coal-biomass serial gasification sub-model. Combustion sub-model includes coal pyrolysis, dense phase combustion, and dilute phase combustion model. Gasification sub-model includes biomass pyrolysis, dense phase coal gasification, dense phase biomass gasification, and dilute phase gasification model. The model studies the effects of key parameters on gasification properties, including gasification temperature, S/B, B/C, and predicts the composition of product gas and gas calorific value along the reactor's axis at different time. The model predictions agree well with experimental results and can be used to study and optimize the operation of the process.