Wen-Yang Chang’s research while affiliated with RiboBio Co. Ltd. and other places

... Mathematics 2024, 12, 2404 2 of 22 feature learning prowess of machine learning models, a multitude of researchers have deployed algorithms like the BP neural network, support vector machine, and random forest for tool RUL prediction [8][9][10]. While these methods have demonstrated effectiveness, they frequently fail to adequately capture the temporal dependencies critical in extensive and complex industrial datasets. ...

Reference:

A Method for Predicting Tool Remaining Useful Life: Utilizing BiLSTM Optimized by an Enhanced NGO Algorithm

... We concatenate these 10 vectors vertically to constituting the matrix presented as PFH p in 10 , ϕ i,10 , θ i,10 ] ∈ R 10×3 . Next, we refer to the Fast PFH (FPFH) [36] descriptor to improve computational efficiency, as follows: ...

Reference:

PosE-Enhanced Point Transformer with Local Surface Features (LSF) for Wood–Leaf Separation

... An IoT platform for intelligent chatter suppression was conducted by Chang et. al. [34]. The system could collect the cutting data, analyze the vibration, and upload it to the cloud for remote monitoring with reference to the stability lobe diagram. ...

Reference:

Reliable Deep Learning and IoT-Based Monitoring System for Secure Computer Numerical Control Machines Against Cyber-Attacks With Experimental Verification

... AI applications in machining, according to the published literature so far, are mostly focused on using a single evaluation criterion, such as surface roughness [5][6][7] or tool wear [8][9][10], for process assessment and optimisation. This is even though machining is a very complex operation covering several processes defined by the interaction between machine tool, environment, and workpiece. ...

Reference:

AI-based optimisation of total machining performance: A review

... This stage required a training process based on the created dataset of the start, target, and obstacle shapes and their morphologies. In the Sensors 2022, 22, 1697 3 of 17 next step, the Q-learning algorithm is used to determine the optimal actions in a gridded 3D workspace so that a robot arm could begin traveling from the start cell to the target cell without collision with an obstacle. In a Q-learning algorithm, states are represented by cells in a 3D workspace, and forward, backward, right, left, down, and up are defined as actions of the robot arms. ...

Reference:

Computer Vision-Based Path Planning for Robot Arms in Three-Dimensional Workspaces Using Q-Learning and Neural Networks

... Table 6 provides a breakdown of the main monitoring objects and the proportion of related research literature in the past 5 years, along with the types of signals collected during online monitoring. [330][331][332][333][334][335] Flutter and processing deformation monitoring -Acoustic emission -Cutting force -Spindle current, acoustic emission -Vibration displacement -Vibration acceleration [336][337][338][339][340][341][342] Surface integrity monitoring -Cutting force [343][344] Roughness monitoring -Acoustic emission -Cutting force [345][346][347][348][349] From Table 6, it is evident that the cutting force signal is the primary signal used for monitoring the machining status. Researchers widely agree that cutting force is highly effective in assessing the condition of tool and workpiece during the machining process. ...

Reference:

Overcoming challenges: advancements in cutting techniques for high strength-toughness alloys in aero-engines

... In the early days of machining, contact-based solutions were the only available methods and those are still a popular choice, even in combination with modern production technologies like additive manufacturing [6]. Non-contact methods based on vision systems [7] or 3D measurements [2,13,16] are popular because of their versatility, speed and high accuracy. ...

Reference:

Multi-operation blank localization with hybrid point cloud and feature-based representation

... In order to reduce inventory pressure, companies need to accurately predict the volume of orders from customers, so that the company's own competitiveness and profits are better than other competitors. How to establish a production predicting mechanism to become a useful competitive among companies [1][2][3]. ...

Reference:

Using Artificial Intelligence in IC Substrate Production Predicting

... During the next phase, we should update the networks regularly. For each time step, if the time step is in the update period, we sample a random minibatch V without replacement from the replay buffer D. In the successive policy decision process, in order to solve the correlation problem between Algorithm 1: Proposed situation-aware autonomous nonlinear drone mobility control algorithm 1 Initialize the critic and actor networks of the drone agent with weights θ Q and θ µ 2 Initialize the target networks as: θQ ← θ Q , θμ ← θ µ 3 for episode = 1, E do 4 Phase 1. Initialize the replay buffer D 5 for mini batch = 1 to c do 6 Randomly generate ϕ states x ∈ X 7 Observe surrounding obstacles through Raycast and store to the states x to aware situation 8 Get corresponding a set of actions u = µ(x|θμ) ∈ U for each x 9 Input the state-action pairs to predefined drone environments and get a set of reward r for each pair, and observe the next set of states x ∈ X 10 end 11 Store the transition pairs ξ = (x, u, r, x ) as a minibatch, which composes D. 12 end 13 Phase 2. Update neural networks periodically 14 for time step = 1, T do 15 If time step is update period, do followings: 16 Sample a random minibatch V without replacement from D 17 Set y i = r i + γQ(x i , µ(x i |θμ)|θQ) 18 Update the θ Q by applying stochastic gradient descent to the loss function of critic network, which can be obtained as ...

Reference:

Situation-Aware Deep Reinforcement Learning for Autonomous Nonlinear Mobility Control in Cyber-Physical Loitering Munition Systems

... The wear types of the turning tool mostly include the flank (edge) wear, crater (face) wear, and nose wear that depended on cutting conditions and the properties of workpiece and tool materials [30][31][32][33][34][35]. In this study, the turning tool life is estimated using flank wear that is characterized by wear height of wear band, as shown in Fig. 2a. ...

Reference:

Investigated iterative convergences of neural network for prediction turning tool wear