Power System Flow Adjustment and Sample Generation Based on Deep Reinforcement Learning

With the increasing complexity of power system structures and the increasing penetration of renewable energy,the number of possible power system operation modes increases dramatically.It is difficult to make manual power flow adjustments to establish an initial convergent power flow that is suitable for operation mode analysis.At present,problems of low efficiency and long time consumption are encountered in the formulation of operation modes,resulting in a very limited number of generated operation modes.In this paper,we propose an intelligent power flow adjustment and generation model based on a deep network and reinforcement learning.First,a discriminator is trained to judge the power flow convergence,and the output of this discriminator is used to construct a value function.Then,the reinforcement learning method is adopted to learn a strategy for power flow convergence adjustment.Finally,a large number of convergent power flow samples are generated using the learned adjustment strategy.Compared with the traditional flow adjustment method,the proposed method has significant advantages that the learning of the power flow adjustment strategy does not depend on the parameters of the power system model.Therefore,this strategy can be automatically learned without manual intervention,which allows a large number of different operation modes to be efficiently formulated.The verification results of a case study show that the proposed method can independently learn a power flow adjustment strategy and generate various convergent power flows.

Numerical Simulation of Axial Inflow Characteristics and Aerodynamic Noise in a Large-Scale Adjustable-Blade Fan

Numerical simulation are conducted to explore the characteristics of the axial inflow and related aerodynamic noise for a large-scale adjustable fan with the installation angle changing from−12°to 12°.In such a range the maximum static(gauge)pressure at the inlet changes from−2280 Pa to 382 Pa,and the minimum static pressure decreases from−3389 Pa to−8000 Pa.As for the axial intermediate flow surface,one low pressure zone is located at the junction of the suction surface and the hub,another is located at the suction surface close to the casing position.At the outlet boundary,the low pressure is negative and decreases from−1716 Pa to−4589 Pa.The sound pressure level of the inlet and outlet noise tends to increase monotonously by 11.6 dB and 7.3 dB,respectively.The acoustic energy of discrete noise is always higher than that of broadband noise regardless of whether the inlet or outlet flow surfaces are considered.The acoustic energy ratio of discrete noise at the inlet tends to increase from 0.78 to 0.93,while at the outlet it first decreases from 0.79 to 0.73 and then increases to 0.84.

Selective Area Growth and Characterization of GaN Nanorods Fabricated by Adjusting the Hydrogen Flow Rate and Growth Temperature with Metal Organic Chemical Vapor Deposition

CaN nanorods are successfully fabricated by adjusting the flow rate ratio of hydrogen （H2）/nitrogen （N2） and growth temperature of the selective area growth （SAG） method with metal organic chemical vapor deposition （MOCVD）. The SAG template is obtained by nanospherical-lens photolithography. It is found that increasing the flow rate of 1-12 will change the CaN crystal shape from pyramid to vertical rod, while increasing the growth temperature will reduce the diameters of GaN rods to nanometer scale. Finally the CaN nanorods with smooth lateral surface and relatively good quality are obtained under the condition that the H2：N2 ratio is 1：1 and the growth temperature is 1030℃. The good crystal quality and orientation of GaN nanorods are confirmed by high resolution transmission electron microscopy. The cathodoluminescence spectrum suggests that the crystal and optical quality is also improved with increasing the temperature.

Experimental Analysis of the Flow Characteristics of an Adjustable Critical-Flow Venturi Nozzle

The response of an adjustable critical-flow Venturi nozzle is investigated through a set indoor experiments aimed to determine the related critical flow rate,critical pressure ratio,and discharge coefficient.The effect of a variation in the cone displacement and liquid content on the critical flow characteristics is examined in detail and it is shown that the former can be used to effectively adjust the critical flow rate.The critical pressure ratio of the considered nozzle is above 0.85,and the critical flow control deviation of the gas flow is within±3%.Liquid flow can reduce the gas critical mass flow rate accordingly,especially for the cases with larger liquid volume and lower inlet pressure.The set of results and conclusions provided are intended to support the optimization of steam injection techniques in the context of heavy oil recovery processes.

Numerical investigation of flow with floating vegetation island

Floating vegetation island(FVI)provides an effective way to remove excessive nutrition and pollutants in rivers.The Reynolds stress model(RSM)is employed to investigate the hydrodynamic characteristics induced by varied canopy densities of FVI in an open channel.In longitudinal direction,four regions are subdivided according to the flow development process:upstream adjustment region(LUD),diverging flow region(LDF),shear layer growth region(LSD),and flilly developed region.The increasing canopy density accelerates the flow adjustment in the diverging flow region and shear layer growth region,signaling a shorter distance to reach an equilibrium stage,while LUD keeps a constant.The vertical profiles of the normalized velocity are found to be self-similar downstream of the diverging flow region.In the vertical direction,the streamwise velocity profiles in the mixing layer collapse for all densities and obey the hyperbolic tangent law.Normalized penetration depth into the canopy was fitted as a function of dimensionless canopy density given by δc/hc=0.404(CDahc)^-0.316.This finding indicates a large space for rapid water renewal between the canopy region and the underlying water driven by the shear-scale vortices.In the lateral direction,the intensification of secondary current and the increasing number of secondary current cells with increasing canopy density reveal that dense floating canopies contribute to strong momentum exchange.The centers of vortices move as canopy density increases,while the vortices in canopy region do not merge with those in the gap region,as limited by the height and width of the canopy region.The distribution of longitudinal velocity in the transects is significantly influenced by secondary current.