Search

Traditional x-ray sources used today for multiple applications, such as medical imaging (computed tomography, radiography, mammography, and interventional radiology) or industrial inspection, are vacuum based electron beam devices that include several key components, such as electron emitters, electron guns/cathodes, and anodes/targets. The associated electronics for electron beam generation, focusing and control, and beam acceleration are located outside the vacuum chamber. The general topology of these tubes has been directionally unchanged for more than 100 years; however, tube design remains a long, inefficient, tedious, and complex process; blind design of experiments do not necessarily make the process more efficient. As a case study, in this paper, we introduce the differential evolution (DE), an artificial intelligence-based optimization algorithm, for the design optimization of x-ray source beam optics. Using a small-scale design problem, we demonstrate that DE can be an effective optimization method for x-ray source beam optics design. [ABSTRACT FROM AUTHOR]

A coil system to generate a uniform field is urgently needed in quantum experiments. However, general coil configurations based on the analytical method have not considered practical restrictions, such as the region for coil placement due to holes in the center of the magnetic shield, which could not be directly applied in most of the quantum experiments. In this paper, we develop a coil design method for quantum experiments using hybrid machine learning. The algorithm part consists of a machine learner based on an artificial neural network and a differential evolution (DE) learner. The cooperation of both learners demonstrates its higher efficiency than a single DE learner and robustness in the coil optimization problem compared with analytical proposals. With the help of a DE learner, in numerical simulation, a machine learner can successfully design coaxial coil systems that generate fields whose relative inhomogeneity in a 25 mm-long central region is ∼10−6 under constraints. In addition, for experiments, a coil system with 0.069% inhomogeneity of the field, designed by a machine learner, is constructed, which is mainly limited by machining the precision of the circuit board. Benefitting from machine learning's high-dimension optimization capabilities, our coil design method is convenient and has potential for various quantum experiments. [ABSTRACT FROM AUTHOR]

Obtaining the temporal shape of an ultrashort laser pulse using the method of dispersion scan entails solving a nonlinear inverse problem, a challenging prospect on its own, yet still aggravated when the pulse shape being measured is temporally varying from pulse to pulse. For this purpose, we use a Differential Evolution (DE) algorithm enhanced by three different regularization methods. The DE algorithm in its standard form is insufficient for reconstructing the pulse in the case of unstable pulse trains. By modifying it to retrieve two independent functions and with the help of regularization, we were able to show that it is possible to simultaneously infer the average length and the coherence length of the pulses. The latter is the shortest pulse the laser source can produce. We also discuss the three different approaches for regularization used in this paper, and from the numerical results we present, we can conclude that a spline-based regularization method is far superior compared to the two other methods under investigation. [ABSTRACT FROM AUTHOR]

For wideband receiver systems, it is challenging to compensate the in-phase/quadrature (I/Q) phase mismatch by traditional methods, especially with a time delay deviation (TDD) between the I/Q channels. Considering the above situation, this paper proposes a full-scale I/Q phase imbalance model concerning TDD. The model divides phase mismatch into two parts, i.e., the linear phase (LP) part and the nonlinear phase part, and compensates each part with the corresponding compensation module separately. The design strategy of the compensation module is innovatively transformed into a constrained nonlinear optimization problem, and a metaheuristic algorithm, the flower pollination algorithm (FPA), is utilized to be the optimizer. The results of the contrast simulation with the LP elimination method show the efficiency of the proposed method. In addition, the superiority of the FPA-based structure is verified by comparing with other metaheuristic algorithms, the artificial bee colony technique, the bat algorithm, and the differential evolution algorithm, in terms of the compensation accuracy, algorithm stability, runtime consumption, and convergence performance. Ultimately, the image rejection ratio improvement on the actual platform after compensation is measured, which validates the proposed compensation structure and the corresponding optimization method practically, and the FPA is still the best choice among the competent optimizers. [ABSTRACT FROM AUTHOR]

The assembly error of flexible joints and the change in joint stiffness during movement make the actual value of joint parameters inconsistent with the given value, which affects the joint control accuracy. In order to suppress the influence of parameters error, a parameters identification method for flexible joint combined offline identification and online compensation is proposed. First, the offline identification model of inertia, mass, and damping and the online identification model of joint stiffness are established, respectively. Then, a hybrid tracking differentiator based on an improved Sigmoid function is designed to track the differential signals of joint motion parameters, and the Lyapunov function is designed to prove its convergence. The adaptive differential evolution is used as the identification algorithm, and the improved adaptive crossover, mutation factor, and Metropolis acceptance criterion are designed to improve the convergence speed. Finally, a feedforward control structure based on identification is designed to compensate for the model deviation. Simulation and experimental results show that the improved differentiator can effectively improve the tracking speed and derivation accuracy of the signals. Compared with other algorithms, the proposed identification method has a faster convergence speed and higher identification accuracy, and feedforward compensation control can effectively correct model parameters and improve control accuracy. [ABSTRACT FROM AUTHOR]

Proton imaging is a powerful technique for imaging electromagnetic fields within an experimental volume, in which spatial variations in proton fluence are a result of deflections to proton trajectories due to interaction with the fields. When deflections are large, proton trajectories can overlap, and this nonlinearity creates regions of greatly increased proton fluence on the image, known as caustics. The formation of caustics has been a persistent barrier to reconstructing the underlying fields from proton images. We have developed a new method for reconstructing the path-integrated magnetic fields, which begins to address the problem posed by caustics. Our method uses multiple proton images of the same object, each image at a different energy, to fill in the information gaps and provide some uniqueness when reconstructing caustic features. We use a differential evolution algorithm to iteratively estimate the underlying deflection function, which accurately reproduces the observed proton fluence at multiple proton energies simultaneously. We test this reconstruction method using synthetic proton images generated for three different, cylindrically symmetric field geometries at various field amplitudes and levels of proton statistics and present reconstruction results from a set of experimental images. The method we propose requires no assumption of deflection linearity and can reliably solve for fields underlying linear, nonlinear, and caustic proton image features for the selected geometries and is shown to be fairly robust to noise in the input proton intensity. [ABSTRACT FROM AUTHOR]

A novel algorithm for the ultrashort laser pulse characterization method of interferometric frequencyresolved optical gating (iFROG) is presented. Based on a genetic method, namely, differential evolution, the algorithm can exploit all available information of an iFROG measurement to retrieve the complex electric field of a pulse. The retrieval is subjected to a series of numerical tests to prove the robustness of the algorithm against experimental artifacts and noise. These tests show that the integrated error-correction mechanisms of the iFROG method can be successfully used to remove the effect from timing errors and spectrally varying efficiency in the detection. Moreover, the accuracy and noise resilience of the new algorithm are shown to outperform retrieval based on the generalized projections algorithm, which is widely used as the standard method in FROG retrieval. The differential evolution algorithm is further validated with experimental data, measured with unamplified three-cycle pulses from a mode-locked Ti:sapphire laser. Additionally introducing group delay dispersion in the beam path, the retrieval results show excellent agreement with independent measurements with a commercial pulse measurement device based on spectral phase interferometry for direct electric-field retrieval. Further experimental tests with strongly attenuated pulses indicate resilience of differential-evolution-based retrieval against massive measurement noise. [ABSTRACT FROM AUTHOR]