Training neural network to recognize targets needs a lot of samples.People usually get these samples in a non-systematic way,which can miss or overemphasize some target information.To improve this situation,a new meth...Training neural network to recognize targets needs a lot of samples.People usually get these samples in a non-systematic way,which can miss or overemphasize some target information.To improve this situation,a new method based on virtual model and invariant moments was proposed to generate training samples.The method was composed of the following steps：use computer and simulation software to build target object＇s virtual model and then simulate the environment,light condition,camera parameter,etc.;rotate the model by spin and nutation of inclination to get the image sequence by virtual camera;preprocess each image and transfer them into binary image;calculate the invariant moments for each image and get a vectors＇ sequence.The vectors＇ sequence which was proved to be complete became the training samples together with the target outputs.The simulated results showed that the proposed method could be used to recognize the real targets and improve the accuracy of target recognition effectively when the sampling interval was short enough and the circumstance simulation was close enough.展开更多
Life evaluation for newly developed lithium-ion batteries is often constrained by the time-intensive and costly nature of battery testing.This is particularly true in the aerospace industry,where limited comprehensive...Life evaluation for newly developed lithium-ion batteries is often constrained by the time-intensive and costly nature of battery testing.This is particularly true in the aerospace industry,where limited comprehensive data availability significantly hampers life evaluations.Data collected from batteries under diverse operating con-ditions and cell mechanisms provides valuable insights for constructing degradation models.Nevertheless,the nonlinearity in battery degradation across operating conditions,combined with data distribution discrepancies among different cell mechanisms,presents significant challenges in developing degradation models for newly designed batteries.In this study,a stress-informed transfer learning methodology is proposed to accelerate the life evaluation process.Firstly,a stochastic model is employed to capture the nonlinear dynamics inherent in battery degradation under diverse operating conditions.Model migration is implemented to adapt stochastic models to unique degradation trends,ensuring precision under varying stresses.Secondly,a Transformer-based model is developed to accommodate variations in data distributions across different cell mechanisms.Domain-adaptive fine-tuning with specified loss function is then incorporated to address the challenge of limited target degradation features.Finally,a hybrid model is devised by integrating these foundational components,realizing accelerated life evaluation through the utilization of multi-modal data.Experimental results demonstrate that the proposed methodology achieves improvements of 63.40%in MAE and 58.55%in RMSE with 30%training data length compared to mainstream benchmark methods.This highlights the method’s potential as an early-stage screening and assessment tool for newly developed space lithium-ion batteries,complementing conventional cycle life evaluation protocols with accelerated evaluations from limited degradation data.展开更多
基金Supported by the Ministerial Level Research Foundation(404040401)
文摘Training neural network to recognize targets needs a lot of samples.People usually get these samples in a non-systematic way,which can miss or overemphasize some target information.To improve this situation,a new method based on virtual model and invariant moments was proposed to generate training samples.The method was composed of the following steps：use computer and simulation software to build target object＇s virtual model and then simulate the environment,light condition,camera parameter,etc.;rotate the model by spin and nutation of inclination to get the image sequence by virtual camera;preprocess each image and transfer them into binary image;calculate the invariant moments for each image and get a vectors＇ sequence.The vectors＇ sequence which was proved to be complete became the training samples together with the target outputs.The simulated results showed that the proposed method could be used to recognize the real targets and improve the accuracy of target recognition effectively when the sampling interval was short enough and the circumstance simulation was close enough.
基金supported by the National Natural Science Foundation of China under Grant No 62201177,No 62371168Natural Science Foundation of Heilongjiang Province under Grant No YQ2023F006.
文摘Life evaluation for newly developed lithium-ion batteries is often constrained by the time-intensive and costly nature of battery testing.This is particularly true in the aerospace industry,where limited comprehensive data availability significantly hampers life evaluations.Data collected from batteries under diverse operating con-ditions and cell mechanisms provides valuable insights for constructing degradation models.Nevertheless,the nonlinearity in battery degradation across operating conditions,combined with data distribution discrepancies among different cell mechanisms,presents significant challenges in developing degradation models for newly designed batteries.In this study,a stress-informed transfer learning methodology is proposed to accelerate the life evaluation process.Firstly,a stochastic model is employed to capture the nonlinear dynamics inherent in battery degradation under diverse operating conditions.Model migration is implemented to adapt stochastic models to unique degradation trends,ensuring precision under varying stresses.Secondly,a Transformer-based model is developed to accommodate variations in data distributions across different cell mechanisms.Domain-adaptive fine-tuning with specified loss function is then incorporated to address the challenge of limited target degradation features.Finally,a hybrid model is devised by integrating these foundational components,realizing accelerated life evaluation through the utilization of multi-modal data.Experimental results demonstrate that the proposed methodology achieves improvements of 63.40%in MAE and 58.55%in RMSE with 30%training data length compared to mainstream benchmark methods.This highlights the method’s potential as an early-stage screening and assessment tool for newly developed space lithium-ion batteries,complementing conventional cycle life evaluation protocols with accelerated evaluations from limited degradation data.
