Soochow University

Everbright Institute of Semiconductor Photonics Co., Ltd.

Suzhou Everbright Photonics Co., Ltd.

Suzhou Fusion Tech Co., Ltd.

Suzhou Hengrui Disheng Medical Co., Ltd.

Inspur Suzhou Intelligent Technology Co., Ltd

Pgytech Co., Ltd.

Deartech Machine Tool (suzhou) Corp., Ltd.

Jun Wang

Lei Zhang

Zhongkui Zhu

Weiguo Huang

Changqing Shen

Xingxing Jiang

Juanjuan Shi

Dong Wang

Dongmiao Song

Shaoyang Tan

Aiwen Zhang

Xu Wang

Jianqin Zhou

Li Zhou

Quanling Li

Guifu Du

Dayong Min

Guofeng Kan

Yongjun Shen

Pengcheng Long

Jun Zhu

Yu Xia

Liang Chen

Xiuzhi Yu

Hao Yu

Bo Li

Jun Dai

Wangping Long

Wenjun Guo

Heng Liu

Huadong Pan

Yao Xiao

Shujuan Sun

Hua Feng

Shuangjie Liu

Zaigang Chen

Xiaofeng Shang

Jianzhe Li

Guangshan Wang

Lin Kong

Nan Chen

Jing Xie

Hanshen Wang

Yiwen Hu

Xinsheng Liao

Yintao Guo

Chuancang Ding

Bangguo Wang

Lichen Zhang

Jingwei Hu

Jinjing Zhang

Yang Cheng

Piao Lei

Yudan Gou

Xiaoyu Wu

Minjie Chen

Hongbo Que

Wu Zhao

Lixin Wang

Zhenguang Li

A semiconductor structure and a method for manufacturing a semiconductor structure are provided. The semiconductor structure includes: a semiconductor substrate layer; an N-type waveguide structure arranged on the semiconductor substrate layer; and an active layer arranged on a surface of the N-type waveguide structure on a side away from the semiconductor substrate layer. The N-type waveguide structure includes a first N-type waveguide layer and a second N-type waveguide layer that are stacked. The second N-type waveguide layer is arranged between the first N-type waveguide layer and the active layer. A conduction band level of the first N-type waveguide layer is the same as a conduction band level of the second N-type waveguide layer. A valence band level of the first N-type waveguide layer is lower than a valence band level of the second N-type waveguide layer. The semiconductor structure increases light emitting efficiency.

Semiconductor structure and method for manufacturing same

A multi-active area semiconductor structure and a method for manufacturing same. The multi-active area semiconductor structure includes: a (2k−1)common confining layer arranged between a kactive layer and a ktunnel junction and in contact with the ktunnel junction; and a 2kcommon confining layer arranged between the kactive layer and a (k+1)tunnel junction and in contact with the ktunnel junction, where a forbidden band width of a kquantum well layer is less than both a forbidden band width of a kfirst-semiconductor layer and a forbidden band width of a ksecond-semiconductor layer; a total thickness of the (2k−1)common confining layer and the 2kcommon confining layer is greater than a critical optical field coupling thickness and less than or equal to twice the critical optical field coupling thickness; and a thickness of the kquantum well layer is less than or equal to 1/10 of a thickness of the (2k−1)common confining layer, and the thickness of the kquantum well layer is less than or equal to 1/10 of a thickness of the 2kcommon confining layer. The multi-active area semiconductor structure has effectively improved light-emission efficiency and reduced optical field crosstalk.

Multi-active area semiconductor structure and method for manufacturing same

The present invention provides an unsupervised fault diagnosis method for mechanical equipment based on an adversarial flow model, main steps including: data preprocessing, converting a mechanical vibration signal into a frequency domain signal, and normalizing the amplitude value of the signal into a range of [0, 1]; prior distribution designing: designing a mixture of Gaussian distribution with K subdistributions, wherein K is determined by the number of mechanical equipment status; model construction: constructing an unsupervised fault diagnosis model by combining an autoencoder, a flow model, and a classifier; model training: training the unsupervised fault diagnosis model by using various classes of status data, along with the designed prior distribution, preset training steps, loss functions, and an optimization algorithm; and fault diagnosis: inputting status data of mechanical equipment into the trained unsupervised fault diagnosis model to obtain a data clustering result and a fault diagnosis result.

Unsupervised fault diagnosis method for mechanical equipment based on adversarial flow model

The present invention provides a dynamic joint distribution alignment network-based bearing fault diagnosis method under variable working conditions, including acquiring bearing vibration data under different working conditions to obtain a source domain sample and a target domain sample; establishing a deep convolutional neural network model with dynamic joint distribution alignment; feeding both the source domain sample and the target domain sample into the deep convolutional neural network model with initialized parameters, and extracting, by a feature extractor, high-level features of the source domain sample and the target domain sample; calculating a marginal distribution distance and a conditional distribution distance; obtaining a joint distribution distance according to the marginal distribution distance and the conditional distribution distance, and combining the joint distribution distance and a label loss to obtain a target function; and optimizing the target function by using SGD, and training the deep convolutional neural network model.

Dynamic joint distribution alignment network-based bearing fault diagnosis method under variable working conditions

The present invention discloses a method for predicting bearing life based on a hidden Markov model (HMM) and transfer learning, including the following steps: (1) acquiring an original signal of full life of a rolling bearing; and extracting a feature set including a time domain feature, a time-frequency domain feature, and a trigonometric function feature; (2) inputting the feature set into an HMM to predict a hidden state, to obtain a failure occurrence time (FOT); (3) constructing a multilayer perceptron (MLP) model, obtaining a domain invariant feature and an optimal model parameter, and obtaining a neural network life prediction model; and (4) inputting the remaining target domain feature sets into the neural network life prediction model, and predicting the remaining life of the bearing. In the present invention, MLP-based transfer learning is used to resolve distribution differences in a source domain and a target domain caused by different operating conditions.

Method for predicting bearing life based on hidden Markov model and transfer learning

A data transmission method, apparatus, and device and a storage medium. A control instruction sent by a first node main program is received and analyzed; when the control instruction is analyzed to be a first control instruction for controlling a first node, a value is assigned to a first variable of the first node; the first control instruction is executed to obtain first result data and fed back same to the first node main program; the first variable of the first node is sent to a second node SES program, so that the second node SES program receives and analyzes same, and assigns a value to a second variable of a second node. According to the method, in a multi-controller data processing situation, node data transmission is increased, redundant link control of a two-node SES program by one node is implemented, and an information interaction redundancy mechanism can be enhanced.

Data transmission method, apparatus, and device, and storage medium

The present disclosure relates to a support frame including a first pan-tilt connector and a leg assembly connected thereto and including a primary leg and a secondary leg. A leg adjustment structure is provided between and configured to unlock or lock the primary leg and the secondary leg, so that the primary leg and the secondary leg can move relative to each other or keep relatively stationary; the secondary leg comprises a first sub-leg and a second sub-leg which are configured such that the first sub-leg and the second sub-leg can be unfolded or folded with respect to each other; when the first sub-leg and the second sub-leg are unfolded, the primary leg, the first sub-leg and the second sub-leg are in a three-leg support state. The present disclosure further provides a photographing device, which comprises the support frame and an image acquisition apparatus mounted on the support frame.

Support frame and photographing device

The invention relates to a fault diagnosis method for a rolling bearing under variable working conditions. Based on a convolutional neural network, a transfer learning algorithm is combined to handle the problem of the reduced universality of deep learning models. Data acquired under different working conditions is segmented to obtain samples. The samples are preprocessed by using FFT. Low-level features of the samples are extracted by using improved ResNet-50, and a multi-scale feature extractor analyzes the low-level features to obtain high-level features as inputs of a classifier. In a training process, high-level features of training samples and test samples are extracted, and a conditional distribution distance between them is calculated as a part of a target function for backpropagation to implement intra-class adaptation, thereby reducing the impact of domain shift, to enable a deep learning model to better carry out fault diagnosis tasks.

Intra-class adaptation fault diagnosis method for bearing under variable working conditions

The invention provides an adaptive manifold probability distribution-based bearing fault diagnosis method, including constructing transferable domains and transfer tasks; converting a data sample in each transfer task into frequency domain data via Fourier transform, inputting the frequency domain data into a GFK algorithm model, and calculating a manifold feature representation matrix related to a bearing fault in each transfer task by using the GFK algorithm model; calculating a cosine distance between centers of a target domain and a source domain in each transfer task according to a manifold feature representation, and defining a target function of in-domain classifier learning; then solving the target function, to obtain a probability distribution matrix of the target domain; and selecting a label corresponding to the largest probability value corresponding to each data sample in the target domain from the probability distribution matrix as a predicted label of the data sample in the target domain.

Adaptive manifold probability distribution-based bearing fault diagnosis method

The present invention discloses a fault diagnosis method under a convergence trend of a center frequency, including: (1) acquiring a dynamic signal x(t) of a rotary machine equipment; (2) setting initial decomposition parameters of a variational model; (3) decomposing the dynamic signal x(t) by using the variational model with the set initial decomposition parameters, and traversing a signal analysis band and performing iterative decomposition on the dynamic signal x(t) under the guidance of a convergence trend of a center frequency, to obtain optimized modals {m. . . m. . . m} and corresponding center frequencies {&#x3c9;. . . &#x3c9;. . . &#x3c9;}; (4) searching a fault related modal m, guiding parameter optimization by using a center frequency &#x3c9;of the fault related modal m, and retrieving an optimal target componentincluding fault information; and (5) performing envelopment analysis on the optimal target component, and diagnosing the rotary machine equipment according to an envelope spectrum.

Fault diagnosis method under convergence trend of center frequency

The present application relates to the technical field of semiconductor optoelectronics, in particular to a multi-active-region cascaded semiconductor laser. The multi-active-region cascaded semiconductor laser comprises: a plurality of cascaded active regions, wherein each cascaded active region comprises a plurality of active regions; and a tunnel junction, arranged on at least one side of the cascaded active region and electrically connected with the cascaded active region; wherein in the cascaded active region, at least one group of adjacent active regions are connected through a barrier layer. In this way, more active regions are added in the periodic gain structure, which improves the internal quantum efficiency of the device and also reduces the carrier density, thereby obtaining more gains. The barrier layer connection does not have the property of introducing a new pn junction, so the layer will not increase the turn-on voltage for device operation, and meanwhile the epitaxial growth is much simpler than that of the tunnel junction.

Multi-active-region cascaded semiconductor laser

A beam combining device and method for a Bragg grating external-cavity laser module has a plurality of side by side light-emitting modules that use a Bragg grating to perform wavelength locking. Output light of the modules is incident to a beam combining element after passing through a focusing optical element for beam combining, and light subjected to beam combining is reflected partially and transmitted partially under the effect of a light splitting element. A part is incident into a dispersion element at a diffraction angle of the element. Parallel light is formed under the effect of a conversion optical element. Spots of the light beams of corresponding wavelengths of the light-emitting modules are formed on an image acquisition mechanism. Whether the wavelengths of the corresponding light-emitting modules are locked is determined by whether there is a deviation between preset spots and spots formed by the module on the acquisition mechanism.

Beam combining device and beam combining method for Bragg grating external-cavity laser module

3D printing methods for workpiece supports, support structures, and workpieces having supports are disclosed. In an embodiment, a printing method of a workpiece support includes the following steps. (1) Configuring a first printing scheme by a printing software installed in a printing apparatus and configuring a workpiece support model according to the first printing scheme. (2) Printing a workpiece support skeleton according to the first printing scheme and the workpiece support model by the printing apparatus and obtaining the workpiece support by filling the workpiece support skeleton. Optionally, step (2) includes controlling a second nozzle to eject a ceramic wire according to the first printing scheme and the support model and controlling a first nozzle to eject a linear material according to the support model to fill the workpiece support skeleton.

Printing method for workpiece support, support structure, and workpiece with support

Disclosed are methods and systems for optimizing printing of a ceramic isolation layer. In some embodiments, the method includes the following steps: preparing a workpiece before printing; printing the workpiece by an optimal printing solution, the optimal printing solution satisfying a setting of key data when printing the ceramic isolation layer; and processing the workpiece after printing to obtain a finished workpiece. In other embodiments, the optimal printing solution is determined by the following steps: printing and processing the ceramic isolation layer and the workpiece isolated by the ceramic isolation layer for multiple times; adjusting the key data by determining a strength of the ceramic isolation layer after printing and deformation data of the workpiece; selecting the ceramic isolation layer parameters and the printing parameters; and taking the setting of the key data as the optimal solution when the deformation data reaches a preset threshold.

Method and system of optimizing printing of ceramic isolation layer

A device for rapidly loading and releasing a syringe. A limiting flange and clamping flange are on an outer side wall of an end, connected to the device, of the syringe. The clamping flange is close to a rear end of the syringe. The device includes an injection head. An outer collar, middle plate, and inner collar are sequentially inside the injection head. The outer collar includes an upper and lower part that are disposed up and down. The inner collar includes an upper part of the inner collar and a lower part of the inner collar that are disposed up and down. Elastic structures are between the upper part of the inner collar and injection head and between the lower part of the inner collar and injection head. The injector syringe can be loaded onto the device and taken out rapidly if the syringe is rotated.

Syringe, and device for rapidly loading and releasing the same

Growing community of inventors

Suzhou, China

Jun Wang