摘要 :
The existing roll state estimation method of the tank semi-trailer is usually based on a roll state characterization parameter and the corresponding threshold. Although these methods are easy to implement, their reliability and fa...
展开
The existing roll state estimation method of the tank semi-trailer is usually based on a roll state characterization parameter and the corresponding threshold. Although these methods are easy to implement, their reliability and fault tolerance are difficult to meet the need of practical engineering applications. To address this problem, a novel estimation method is proposed to achieve the accurate roll state of the tank semi-trailer, relying on fusing the current and historical information of multiple vehicle state parameters. First, the roll state estimator is designed based on a Kalman filter and an improved Long Short-Term Memory (LSTM) network. Its inputs are the liquid-filling ratio, roll angle, lateral acceleration and yaw rate. Its output is the lateral-load transfer rate on the last axle wheel
$(\Delta L)$
), which is used to judge the roll state. The high-risk roll state amplification factor is presented in the loss function of the LSTM network to enhance the accuracy of the high-risk roll state estimation. Then, with the sample data acquisition system, typical rollover scene and acquisition process, the real vehicle experiment is carried out in the closed field to collect the sample data of the LSTM network inputs and output. During the experiment, the true value of
$\Delta L$
can be directly calculated by the wheel force sensor data. Finally, comparing
$\Delta L$
output by the estimator with the true value of
$\Delta L$
, it is validated that the proposed method can provide the reliable and fault-tolerant estimation of the tank semi-trailer roll state.
收起
摘要 :
The existing roll state estimation method of the tank semi-trailer is usually based on a roll state characterization parameter and the corresponding threshold. Although these methods are easy to implement, their reliability and fa...
展开
The existing roll state estimation method of the tank semi-trailer is usually based on a roll state characterization parameter and the corresponding threshold. Although these methods are easy to implement, their reliability and fault tolerance are difficult to meet the need of practical engineering applications. To address this problem, a novel estimation method is proposed to achieve the accurate roll state of the tank semi-trailer, relying on fusing the current and historical information of multiple vehicle state parameters. First, the roll state estimator is designed based on a Kalman filter and an improved Long Short-Term Memory (LSTM) network. Its inputs are the liquid-filling ratio, roll angle, lateral acceleration and yaw rate. Its output is the lateral-load transfer rate on the last axle wheel
$(\Delta L)$
), which is used to judge the roll state. The high-risk roll state amplification factor is presented in the loss function of the LSTM network to enhance the accuracy of the high-risk roll state estimation. Then, with the sample data acquisition system, typical rollover scene and acquisition process, the real vehicle experiment is carried out in the closed field to collect the sample data of the LSTM network inputs and output. During the experiment, the true value of
$\Delta L$
can be directly calculated by the wheel force sensor data. Finally, comparing
$\Delta L$
output by the estimator with the true value of
$\Delta L$
, it is validated that the proposed method can provide the reliable and fault-tolerant estimation of the tank semi-trailer roll state.
收起
摘要 :
The existing roll state estimation method of the tank semi-trailer is usually based on a roll state characterization parameter and the corresponding threshold. Although these methods are easy to implement, their reliability and fa...
展开
The existing roll state estimation method of the tank semi-trailer is usually based on a roll state characterization parameter and the corresponding threshold. Although these methods are easy to implement, their reliability and fault tolerance are difficult to meet the need of practical engineering applications. To address this problem, a novel estimation method is proposed to achieve the accurate roll state of the tank semi-trailer, relying on fusing the current and historical information of multiple vehicle state parameters. First, the roll state estimator is designed based on a Kidman filter and an improved Long Short-Term Memory (LSTM) network. Its inputs are the liquid-filling ratio, roll angle, lateral acceleration and yaw rate. Its output is the lateral-load transfer rate on the last axle wheel (ΔL), which is used to judge the roll state. The high-risk roll state amplification factor is presented in the loss function of the LSTM network to enhance the accuracy of the high-risk roll state estimation. Then, with the sample data acquisition system, typical rollover scene and acquisition process, the real vehicle experiment is carried out in the closed Geld to collect the sample data of the LSTM network inputs and output. During the experiment, the true value of ΔL can be directly calculated by the wheel force sensor data. Finally, comparing ΔL output by the estimator with the true value of ΔL., it is validated that the proposed method can provide the reliable and fault-tolerant estimation of the tank semi-trailer roll state.
收起
摘要 :
Flood control operation is the top priority for the management of the Yangtze River.To improve the timeliness and reliability of operation decision,it is of great significance to establish a set of intelligent decision system for ...
展开
Flood control operation is the top priority for the management of the Yangtze River.To improve the timeliness and reliability of operation decision,it is of great significance to establish a set of intelligent decision system for flood control management,which can serve multi-level users.In order to achieve the goal,the key is to modularize the core business functions through the construction of topology relations of river network and control sites for the Yangtze River.In this paper,it would integrate the functions of hydrological forecast,flood propagation,flood control operation and flood risk analysis for the core business.Based on the core business model and information expansion interface,we develop flood control decision support system of the Yangtze River by virtue of GIS visualization and interactive WebService technology.The system can realize the real-time query of water-rainfall information and the fast-time simulation of flood control scheduling scheme,so as to provide the professional,scientific and accurate technical support for the group decision in the flood control.
收起
摘要 :
Personalized customization has become an effective approach
to satisfying the personalized requirements with the
development of the Internet. And the performance of
individual customization system depends on the structure of
p...
展开
Personalized customization has become an effective approach
to satisfying the personalized requirements with the
development of the Internet. And the performance of
individual customization system depends on the structure of
product configuration model. In order to make the customized
product closely meet customer's real individualization and
provide the system with functions of evaluation and
recommendation, a product configuration model based on
customer interest orientation was proposed. Besides, the
manufacturing and fitting time of parts and cost factors are
considered in the model, which makes personalized
requirements of customer be satisfied better. The product
performance is quantified by using concepts of parts
performance index (PPI) and product performance matrix.
The solution to engineering configuration corresponding to
interest orientation is developed through the analytic
hierarchy process(AHP) and matrix theory. Genetic
algorithms are used in the model, and compared with the
exhaustive method and the elimination by contributing
weights method, which are two traditional product
configuration algorithms, a conclusion is drawn that both the
precision and efficiency of calculation are improved
remarkably.
收起
摘要 :
Post-disaster transition is necessary in disaster relief. Care, security and sustainable development are the basic principles of the temporary resettlement area planning. The planning has particularity in the use of function, safe...
展开
Post-disaster transition is necessary in disaster relief. Care, security and sustainable development are the basic principles of the temporary resettlement area planning. The planning has particularity in the use of function, safety precautions, the project cost, psychological relief and so on. Architectural design also has special technical requirements in the use of the function and cost control and sustainable development.
收起
摘要 :
Post-disaster transition is necessary in disaster relief. Care, security and sustainable development are the basic principles of the temporary resettlement area planning. The planning has particularity in the use of function, safe...
展开
Post-disaster transition is necessary in disaster relief. Care, security and sustainable development are the basic principles of the temporary resettlement area planning. The planning has particularity in the use of function, safety precautions, the project cost, psychological relief and so on. Architectural design also has special technical requirements in the use of the function and cost control and sustainable development.
收起
摘要 :
With the development of technology, there is an increasing demand for wireless charging. Among them, the magnetic resonance wireless energy transfer technology is developing particularly rapidly because of its high transmission ef...
展开
With the development of technology, there is an increasing demand for wireless charging. Among them, the magnetic resonance wireless energy transfer technology is developing particularly rapidly because of its high transmission efficiency. However, the impedance matching between devices in a wireless energy transfer system determines the efficiency of the transfer. A good impedance match plays a decisive role in improving the transmission efficiency. In this paper, we use a vector network analyzer to observe the return loss values of devices, design matching circuits and perform impedance matching adjustment.
收起
摘要 :
High-frequency transformers with small size and light weight have been widely concerned by researchers at home and abroad. The oil-paper insulation of high frequency transformers is affected by the high frequency pulse produced by...
展开
High-frequency transformers with small size and light weight have been widely concerned by researchers at home and abroad. The oil-paper insulation of high frequency transformers is affected by the high frequency pulse produced by power electronic device, which is more prone to discharge fault and insulation failure. To study the effect of high frequency pulse voltage on the discharge characteristics of oil-paper insulation, a needle plate electrode model was used to carry out the discharge experiment of different layers of oil-paper insulation samples at 1–20 kHz bipolar high frequency square wave voltage produced by high frequency power supply. The partial discharge and breakdown characteristics of oil-paper insulation under high frequency square wave pulse voltage were studied. The results show that the discharge quantity of oilpaper insulation increases greatly with high frequency pulse voltage compared with AC voltage. There is no discharge phenomenon in the process of discharge development, which is significantly different from that under AC voltage. The intensity of discharge increases under high frequency pulse voltage and its increasement accelerates the deterioration of oil-paper insulation.
收起
摘要 :
High-frequency transformers with small size and light weight have been widely concerned by researchers at home and abroad. The oil-paper insulation of high frequency transformers is affected by the high frequency pulse produced by...
展开
High-frequency transformers with small size and light weight have been widely concerned by researchers at home and abroad. The oil-paper insulation of high frequency transformers is affected by the high frequency pulse produced by power electronic device, which is more prone to discharge fault and insulation failure. To study the effect of high frequency pulse voltage on the discharge characteristics of oil-paper insulation, a needle plate electrode model was used to carry out the discharge experiment of different layers of oil-paper insulation samples at 1–20 kHz bipolar high frequency square wave voltage produced by high frequency power supply. The partial discharge and breakdown characteristics of oil-paper insulation under high frequency square wave pulse voltage were studied. The results show that the discharge quantity of oilpaper insulation increases greatly with high frequency pulse voltage compared with AC voltage. There is no discharge phenomenon in the process of discharge development, which is significantly different from that under AC voltage. The intensity of discharge increases under high frequency pulse voltage and its increasement accelerates the deterioration of oil-paper insulation.
收起