Haibo Shi

Haibo Shi

Auto-generated profile, based on publicly available information

Haibo Shi

Key Laboratory of Networked Control Systems, Chinese Academy of Sciences, Shenyang, China

Research Focus & Metrics

Academic Career Length

20Years (2007-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 July 2022 in IEEE Transactions on Industrial Informatics
Verified
Journal Article
Closed access
XZXiaofeng Zhou
Xiaofeng Zhou
NZNaiju Zhai
Naiju Zhai
SLShuai Li
Shuai Li
HSHaibo Shi
Haibo Shi
Industrial time series, as a kind of data that responds to production process information, can be analyzed and predicted for effective monitoring of industrial production processes. There are problems of data shortage and algorithm cold start in industrial modeling process caused by complex working conditions, change of data acquisition environment and short running time of equipment. As a result, the accuracy of the existing data-driven industrial time series prediction algorithm is greatly limited. To address the above problems, we propose a new time series prediction method for industrial processes under limited data based on dynamic transfer learning in this work. This method aims to effectively use historical data of similar equipment or working conditions rather than discard them to help establish an industrial time series prediction model with limited target data. In this method, first, historical data are divided into multiple batches, and then a new multi-source transfer learning framework with dynamic maximum mean difference (MMD) loss is established according to the distribution distance between each batch of historical data and the limited target data at the current moment. The framework also combines multi-task learning methods to establish multi-step prediction model for online learning in industrial processes. Compared with other commonly used methods, experiments on two real-world datasets of solar power generation prediction and heating furnace temperature prediction demonstrate the effectiveness of the proposed method.

21 May 2019 in Processes
Verified
Journal Article
Open access
PDF file
ZHZhonghua Han
Zhonghua Han
QZquan zhang
quan zhang
HSHaibo Shi
Haibo Shi
JZJingyuan Zhang
Jingyuan Zhang
Flow shop scheduling optimization is one important topic of applying artificial intelligence to modern bus manufacture. The scheduling method is essential for the production efficiency and thus the economic profit. In this paper, we investigate the scheduling problems in a flexible flow shop with setup times. Particularly, the practical constraints of the multi-queue limited buffer are considered in the proposed model. To solve the complex optimization problem, we propose an improved compact genetic algorithm (ICGA) with local dispatching rules. The global optimization adopts the ICGA, and the capability of the algorithm evaluation is improved by mapping the probability model of the compact genetic algorithm to a new one through the probability density function of the Gaussian distribution. In addition, multiple heuristic rules are used to guide the assignment process. Specifically, the rules include max queue buffer capacity remaining (MQBCR) and shortest setup time (SST), which can improve the local dispatching process for the multi-queue limited buffer. We evaluate our method through the real data from a bus manufacture production line. The results show that the proposed ICGA with local dispatching rules and is very efficient and outperforms other existing methods.

23 October 2024 in IEEE Transactions on Industrial Informatics
Verified
Journal Article
Closed access
MWMengqing Wu
Mengqing Wu
XZXiaofeng Zhou
Xiaofeng Zhou
SLShuai Li
Shuai Li
HSHaibo Shi
Haibo Shi
Deep learning models have gained significant attention and application in recent years to improve the accuracy and efficiency of industrial time series prediction. However, the dynamic changes in industrial processes present a key challenge for data-driven models. Specifically, the performance of deployed models deteriorates over time and fails to adapt to new operating conditions. Currently, two common update methods exist: Retraining the model using historical and new operating data, which incurs high computation and storage costs, or incrementally fine-tuning the model solely using new data, which leads to catastrophic forgetting of learned patterns. To address these issues, this article proposes an adaptive continual learning method for nonstationary industrial time series prediction. Our approach tackles the problems by hint-based network parameter learning to retain the dark knowledge from previous tasks and avoid catastrophic forgetting of accumulated knowledge. In addition, we design a soft buffer to aid memory and learning of key patterns under the current operating condition. Lastly, a time-sensitive activation function is proposed to endow the neural network with time-evolving properties, thereby enhancing the model's generalization ability. Compared with other update methods and different continual learning methods, the superiority of our method is validated on solar power generation data and real data of grinding and grading process.

1 October 2019 in Processes
Verified
Journal Article
Open access
PDF file
ZHZhonghua Han
Zhonghua Han
CHChao Han
Chao Han
SLShuo Lin
Shuo Lin
XDXiaoting Dong
Xiaoting Dong
HSHaibo Shi
Haibo Shi
Actual manufacturing enterprises usually solve the production blockage problem by increasing the public buffer. However, the increase of the public buffer makes the flexible flow shop scheduling rather challenging. In order to solve the flexible flow shop scheduling problem with public buffer (FFSP–PB), this study proposes a novel method combining the simulated annealing algorithm-based Hopfield neural network algorithm (SAA–HNN) and local scheduling rules. The SAA–HNN algorithm is used as the global optimization method, and constructs the energy function of FFSP–PB to apply its asymptotically stable characteristic. Due to the limitations, such as small search range and high probability of falling into local extremum, this algorithm introduces the simulated annealing algorithm idea such that the algorithm is able to accept poor fitness solution and further expand its search scope during asymptotic convergence. In the process of local scheduling, considering the transferring time of workpieces moving into and out of public buffer and the manufacturing state of workpieces in the production process, this study designed serval local scheduling rules to control the moving process of the workpieces between the public buffer and the limited buffer between the stages. These local scheduling rules can also be used to reduce the production blockage and improve the efficiency of the workpiece transfer. Evaluated by the groups of simulation schemes with the actual production data of one bus manufacturing enterprise, the proposed method outperforms other methods in terms of searching efficiency and optimization target.

27 April 2021 in IEEE Access
Verified
Journal Article
Open access
YCYang Cao
Yang Cao
HSHaibo Shi
Haibo Shi
Traditionally, process planning and scheduling were performed sequentially, where scheduling depended on the result of process planning. Considering their complementarity, the two functions are more tightly integrated to improve the performance of job shop flexible manufacturing environment. This study proposes an adaptive multi-strategy artificial bee colony (AMSABC) algorithm to solve integrated process planning and scheduling (IPPS) problem. In AMSABC, two search strategies with different characteristics are introduced into employed bees and onlooker bees to take on the responsibility of both exploration and exploitation. The selection probability of each search strategy is dynamically adjusted according to previous experiences. To further improve the exploitation performance of the approach, a problem-specific multi-objective local search has been embedded in the proposed algorithm. Furthermore, AMSABC algorithm presents a unique solution representation where a food source is represented by three discrete vectors, and a well-designed decoding scheme is developed. Next, the corresponding neighborhood structure is adopted that it can directly generate feasible solutions in the search space. The proposed algorithm is tested on the well-known benchmark instances and compared with the state-of-the-art algorithms. Through detail analysis of experimental results, AMSABC algorithm is more beneficial in the quality and efficiency of solution.

17 April 2023 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
NZNaiju Zhai
Naiju Zhai
XZXiaofeng Zhou
Xiaofeng Zhou
SLShuai Li
Shuai Li
HSHaibo Shi
Haibo Shi
Billet heating temperature directly affects the quality of the billet, but the existing technology cannot measure the billet surface temperature. Therefore, we accurately predict the temperature of the furnace by a soft sensor to approximate the billet temperature. Limited by the complexity of the heating process and the lack of computing resources in factories, the existing research pays little attention to or even cannot meet the temperature prediction requirements of multiple heating zones in multiple heating furnaces. To address the above problems, we propose a temperature prediction method for multi-heating furnaces based on transfer learning and knowledge distillation. The method establishes a multisource domain model of a source domain furnace on a cloud platform. Then, the multisource knowledge is transferred to the target domain heating furnace, and the target domain teacher model is established by fine-tuning the source domain model through the target domain data. Next, to efficiently predict the furnace temperature, a shallow multi-task student model is established at the edge server to predict multiple heating zones in the target furnace. Furthermore, a knowledge distillation method for regression prediction is proposed so that the student model can improve the prediction accuracy under the guidance of the cloud teacher model. The effectiveness of the method is verified by experiments on 20 different heating zone datasets in two heating furnaces and two wind turbine datasets. The consistency of multiple experiments shows that this method can not only improve the accuracy by transferring the source domain knowledge but also reduce the size of model parameters by knowledge distillation on the premise of meeting the requirements of prediction error.