Jane Lehr

Characterization and analysis of a pulse forming network based 11 stage Marx system for a high power microwave, plasma and beam physics test stand

Summary form only given. The University of New Mexico Electrical and Computer Engineering Department is developing advanced high power microwave sources and charged particle beam accelerators and pulsed power capabilities. This paper presents the characterization, analysis, implementation and modification of a high power Marx coupled to a low impedance charged particle beam accelerator. The 11 stage Marx is a bipolarity charged 100 kV device and each stage is comprised of the main capacitor and switch as well as an LCR based pulse forming line and a fast rise time peaking switch. A numerical model of the device has been created that includes the pulse forming network and is compared to experiments at various charging voltages. A configuration as well as experimental data for unipolar charging will also be presented. A variable load impedance has been designed that allows for output characterization of the Marx driving different impedances. To match the driver to a low impedance device and conical transmission line has been designed and numerical as well as experimental results of this are also presented. The initial triggering scheme called for electrical high voltage triggering of the first three stages but a design of a laser trigger based on 800 nm laser burst impinging on one of the electrode and its implementation will also be presented along with jitter measurements.

The optimal planning and dynamic operation of distributed generation method based on modified multiobjective optimization in power distribution system

This paper proposes an optimal method to plan and dynamically operate the distributed generation (DG) based on the modified nondominated sorting genetic algorithm II (NSGA-II). First, the uncertainty of load and DG (photovoltaic panels) output are considered. Daily summer load data and ideal photovoltaic (PV) panels output data are taken from a local electrical company in Albuquerque. To model the uncertainty of load, the daily load is randomized to each bus for every time interval (15 minutes), but with their sum equal to the daily load. The uncertainty of PV output is modeled by adding the cloudy index and frequency factor to the ideal PV output. Second, the placement of a DG is defined by voltage sensitivity analysis. The bus with high voltage sensitivity is considered to have priority for the installation of DG. To find the optimal daily operation of DG, a multiobjective problem is formulated that focuses on the minimization of a circuits voltage deviations, active and reactive power losses. To solve the problem, the traditional NSGA II is modified by incorporating a fuzzy logic decision model. The fuzzy logic model selects an optimally compromised solution from the Pareto front by analyzing its weights of voltage deviations, active and reactive power losses. The selected solution includes the optimal outputs for every generator, synchronous compensator, capacitor bank, and PV panels for each time interval. Furthermore, to operate DG optimally and dynamically, the methods computation speed is crucial. To increase the modified NSGA II computation speed, the population initialization space is reduced and the population is selected and saved for the next generation based on load analysis and experiments. The method is tested on the IEEE 14 bus and a local residential circuit. The results on reducing the power losses, voltage deviations, and increasing the algorithm speed demonstrate the effectiveness of this method.

Distribution-level peak load prediction based on Bayesian Additive Regression Trees

Accurate peak load prediction is an important element of the daily planning, operation, and dispatch scheduling in electric utilities. In order to build a better distribution-level peak load prediction reference model for the local utility in Albuquerque, a new nonparametric method, Bayesian Additive Regression Trees (BART), is introduced. First, a detailed analysis of peak load and local weather information from 2012 and 2013 are used. Strong linear relationship is displayed between the peak load and various weather factors during the winter and spring months and nonlinear relationship dominates in the summer and fall months. Next, the BART method is applied with a principled permutation-based inferential variable selection approach. The BART methods prediction accuracy is then compared with the Multiple Linear Regression (MLR), the result we developed when we cooperated with the local utility company, and the Support Vector Machine (SVM). After thoroughly analyzing and testing the methods based on different parameters, the methods are compared with Mean Square Error (MSE), Root Mean Square Error (RMSE), and Normalized Mean Square Error (NMSE) indexes. The BART displays the best prediction accuracy for every index in our case and its uncertain estimate further provides the confidence interval for the peak load prediction, which also has very high accuracy. Thus, the BART provides the local utility with a better reference peak load forecasting model. Last, influential weather and human factors are summarized.

A spark gap model for LTspice and similar circuit simulation software

In the past, the simulation of pulsed power circuits used specifically designed programs such as Sandias SCREAMER, NRLs BERTHA, or the industry developed TL Code. These codes incorporated specialized models for many pulsed power components which captured key manifestations of the physical phenomena. Now, the simulation of pulsed power circuits is likely to be accomplished with the ubiquitous SPICE-based software. Of the available software, LTspice, a program developed by semiconductor manufacturer Linear Technology Corp, has become very prevalent due to its ease of use, continuous improvement, and free availability. However, LTspice - like all SPICE programs - does not include realistic models for key pulsed power circuit devices - including the spark gap switch. While simple switch models do exist in LTspice and other SPICE programs, these can only crudely approximate the behavior of a spark gap. This effort focuses on developing an LTspice circuit model for a gas-filled spark gap switch that is physically realistic while being simple enough as to permit simulations to run in reasonable times on typical personal computers. Efforts are made to create a model that is portable between different SPICE programs with minimal modification.

Investigation of Monolithic Radial Transmission Lines for Z-Pinch

A mathematical expression of the output voltage from a nonuniform transmission line with an arbitrary input pulse was deduced. The first arriving wave, peak power efficiency, and droop of the output voltage were further clarified using analytical method. The transmission characteristics of monolithic radial transmission lines (MRTLs) with different impedance profiles were investigated by 3-D electromagnetic (EM) simulation and it was found that the hyperbolic impedance profile is the best choice for future Z-pinch drivers. The results obtained from 3-D EM simulation are in good agreement with those obtained from the experiments on a scaled-down MRTL.

Modifications to the von Laue statistical distribution of the times to breakdown at a polymer-oil interface

A statistical analysis has been undertaken to determine the statistical and formative times associated with breakdowns along a polymer-oil interface under impulse conditions. Early analysis was based on an assumption that the breakdown data followed the von Laue Distribution. However, it was found that in the Laue plots there were deviations from the expected straight line behavior at short times to breakdown, which may be due to a normal distribution in values of the formative times. In addition, the plots showed multiple straight line sections, which suggested that changes were occurring to the breakdown processes during the experimental run, or that more than one mechanism of breakdown was occurring. Values of the statistical time ts and the formative time tf were determined from the data by making choices on the straight line section to be considered, and ignoring the effects of the normal distribution on the derived values of ts and tf. The present paper is focused on further development of this statistical method, including a rigorous analysis of the experimental data, taking into account the effect that a normal distribution of the formative times has on the derived values of ts and tf. Optimal fits in terms of three parameters: ts, tf, and σf (the standard deviation of the formative time) have been derived using Kolmogorov-Smirnov statistics to quantify the quality of fit. The quality of these fits and the applicability of this approach is discussed.

Analysis of peak power efficiency and droop from a nonuniform transmission line

A mathematical expression of the output voltage from a nonuniform transmission line (NTL) with an arbitrary input pulse was deduced. Due to this mathematical expression, two transmission characteristics of NTLs with linear, exponential and Gaussian impedance profiles were further clarified. The first one is that the peak power efficiencies of NTLs with a half-sine input voltage are quantified as functions of Ψ (the ratio of the output impedance to the input impedance of the NTLs) and Γ (the ratio of the pulse width to the one-way transit time of the NTLs). The second one is that the top of an initially rectangle input voltage pulse falls at the terminal of the NTL and that the ratio of the droop to the top of the output voltage is also quantified as a function of Ψ and Γ.

Experiments of a monolithic radial transmission line

This paper presents the experimental results of a monolithic radial transmission line (MRTL) that may be used in pulsed power generators and microwave devices. The MRTL with a hyperbolic impedance profile is 508 mm in radius, corresponding to a one-way transit time of 15 ns for the electromagnetic wave. In the experiments, up to twenty identical voltage pulses, 10 ns in FWHM and 2 ns in rise-time, were fed into the MRTL through 20 input BNC connectors that are uniformly distributed along the outer circumference of the MRTL. It was found that the amplitude of the voltage from the output BNC connector located in the center of the MRTL is nearly proportional to the total number of the input branches. The effect of the failure modes on the output voltage was investigated. For the MRTL driven by 20 input branches, while the open-circuit or short-circuit even in one input branch considerably decreases the amplitude of the output voltage, the jitter shorter than 2 ns in 3 input branches makes no obvious effect on the output voltage.

Initial characterization of a modular Magnetically Insulated Line Oscillator

Summary form only given. The high power microwave, plasma and beam physics group at the University of New Mexico, Electrical Engineering Department has developed a modular Magnetically Insulated Line Oscillator (MILO) that will function as a test bed for fundamental and applied physics research. This modular MILO is comprised of stages that allow for easy changes to the internal and external geometric structure. We present numerical simulations, particle in cell as well as radiation transport, on several low voltage operating points for one geometric configuration. Initial experimental data showing the electrical operation of the MILO as driven by a fast rise time Marx is presented and compared to numerical results. Results for time resolved radiation measurements due to electron impact are shown as well as their correlation to experimental electrical measurements and numerical simulations of sheath currents. Experimental resonant mode cold test data is shown along with its numerical simulation analogue.