Search

CN-120940587-B - Casting production intelligent monitoring method based on Internet of things

CN120940587BCN 120940587 BCN120940587 BCN 120940587BCN-120940587-B

Abstract

The application provides an intelligent monitoring method for casting production based on the Internet of things, which comprises the steps of determining bubble generation probability and bubble size distribution range if local difference of casting liquid viscosity exceeds preset bubble generation critical viscosity, maintaining current gate opening time sequence and flow distribution if the local difference of casting liquid viscosity does not exceed preset bubble generation critical viscosity, optimizing multi-gate opening time sequence according to the bubble generation probability and the bubble size distribution range, determining opening sequence and gate opening time sequence deviation of each gate to generate an optimized opening time sequence scheme, dynamically adjusting flow distribution proportion of each gate according to the optimized opening time sequence scheme to generate a flow distribution initial scheme, acquiring casting flow state feedback data after executing the flow distribution initial scheme to generate an updated flow distribution scheme, and calculating internal stress distribution caused by filling integrity of a casting and pressure fluctuation of a stagnation area according to the updated flow distribution scheme to obtain an internal defect prediction result.

Inventors

  • CHEN SHUJUAN

Assignees

  • 梅州华和精密工业有限公司

Dates

Publication Date
20260505
Application Date
20250804

Claims (10)

  1. 1. An intelligent casting production monitoring method based on the Internet of things is characterized by comprising the following steps: Preprocessing casting mold geometric data and casting liquid viscosity distribution data, determining a temperature gradient of a stagnation area and a local difference of casting liquid viscosity according to extracted casting liquid flow state parameters, and analyzing geometric constraints of the stagnation area of a casting mold structure to obtain a local gate pressure gradient of each gate area; if the local difference of the viscosity of the casting solution exceeds the preset bubble generation critical viscosity, determining the bubble generation probability and the bubble size distribution range, and if the local difference of the viscosity of the casting solution does not exceed the preset bubble generation critical viscosity, maintaining the current gate opening time sequence and flow distribution; Optimizing a multi-gate opening time sequence according to the bubble generation probability and the bubble size distribution range, determining the opening sequence of each gate and the gate opening time sequence deviation, and generating an optimized opening time sequence scheme; dynamically adjusting the flow distribution proportion of each gate according to the optimized opening time sequence scheme to generate a flow distribution initial scheme; Acquiring casting liquid flow state feedback data after executing the initial flow distribution scheme, and generating an updated flow distribution scheme; Calculating the filling integrity of the casting and the internal stress distribution caused by the pressure fluctuation of the stagnation area according to the updated flow distribution scheme to obtain an internal defect prediction result; And if the bubble size distribution range in the internal defect prediction result exceeds a preset threshold, iteratively adjusting the gate opening time sequence deviation and the flow distribution proportion to obtain optimized casting parameters, and if the bubble size distribution range does not exceed the preset threshold, confirming the casting parameters and outputting a final casting scheme.
  2. 2. The intelligent monitoring method for casting production based on the internet of things according to claim 1, wherein the preprocessing of casting geometric data and casting liquid viscosity distribution data, determining a local difference of a stagnation area temperature gradient and casting liquid viscosity according to extracted casting liquid flow state parameters, and analyzing a geometric constraint of the stagnation area of a casting mold structure to obtain a local gate pressure gradient of each gate area, comprises: The method comprises the steps of collecting temperature field data and casting flow speed data in a casting mould to generate a temperature distribution matrix and a speed distribution matrix, calculating the temperature change rate of an adjacent area according to the temperature distribution matrix to generate a temperature gradient of a stagnation area, identifying position coordinates of the stagnation area according to the speed distribution matrix, extracting a channel cross section and a corner angle from a three-dimensional model of the casting mould according to the position coordinates of the stagnation area to generate a geometric feature vector of the casting mould, generating a casting flow state parameter according to the casting flow speed data and the temperature distribution matrix, constructing a feature matrix according to the geometric feature vector of the casting mould and the casting flow state parameter, processing the feature matrix through a convolutional neural network to generate a local pressure gradient of a pouring gate of each pouring gate area.
  3. 3. The intelligent monitoring method for casting production based on the internet of things according to claim 2, wherein the constructing a feature matrix according to the casting geometric feature vector and the casting flow state parameter, processing the feature matrix through a convolutional neural network, and generating the gate local pressure gradient of each gate region comprises the following steps: generating a cross-sectional area change rate and a corner resistance factor according to the channel cross-sectional area and the corner angle in the casting geometric feature vector, generating a viscosity gradient component according to the viscosity value in the casting liquid flow state parameter, constructing the feature matrix according to the cross-sectional area change rate, the corner resistance factor and the viscosity gradient component, performing convolution operation and pooling treatment on the feature matrix through a convolution neural network to generate a flow resistance coefficient, and generating a gate local pressure gradient of each gate area according to the flow resistance coefficient and the casting liquid inlet pressure.
  4. 4. The intelligent monitoring method for casting production based on the Internet of things is characterized by optimizing a multi-gate opening time sequence according to bubble generation probability and bubble size distribution range, determining opening sequence and gate opening time sequence deviation of each gate, and comprises calculating a bubble volume expected value of each gate position according to the bubble generation probability and the bubble size distribution range, constructing an optimized objective function according to the bubble volume expected value, encoding gate opening time sequences through a genetic algorithm to generate a time sequence scheme population, calculating a casting liquid front end position track in a casting liquid filling process according to the time sequence scheme population, calculating a space superposition volume according to the casting liquid front end position track and the bubble generation probability, generating an adaptation value, iteratively updating the time sequence scheme population according to the adaptation value, generating the opening sequence of each gate, and calculating adjacent gate opening time intervals according to the opening sequence of each gate to generate the opening sequence deviation.
  5. 5. The intelligent monitoring method for casting production based on the Internet of things is characterized by comprising the steps of calculating the opening time interval of each gate according to the opening sequence of each gate, generating the opening time sequence deviation, calculating the opening time interval of each gate according to the opening sequence of each gate, analyzing the speed vector included angle when two casting solutions corresponding to the adjacent gates are combined according to the opening time interval, adjusting the opening time of the gate with the opening time sequence in the adjacent gate according to the speed vector included angle to obtain the adjusted time, and generating the opening time sequence deviation according to the adjusted time and the total filling time.
  6. 6. The intelligent monitoring method for casting production based on the Internet of things is characterized in that the flow distribution proportion of each gate is dynamically adjusted according to an optimized opening time sequence scheme to generate a flow distribution initial scheme, the method comprises the steps of extracting time sequence data of casting liquid speed in an area according to the optimized opening time sequence scheme, determining the fluctuation amplitude of the casting liquid speed, solving the basic flow value of each gate, calculating the sum of flows of the gates which are simultaneously opened according to the basic flow value of each gate and the time interval in the opening time sequence, and determining the flow distribution initial scheme.
  7. 7. The intelligent monitoring method for casting production based on the internet of things according to claim 1, wherein the step of obtaining the casting flow state feedback data after executing the initial flow distribution scheme and generating the updated flow distribution scheme comprises the following steps: Calculating the flow pulsation amplitude of the casting liquid according to the flow state feedback data of the casting liquid, judging the filling rate balance state according to the flow pulsation amplitude of the casting liquid and the propulsion distance difference value of the front position of the casting liquid, and adopting a reinforcement learning algorithm to adjust the flow distribution proportion of each pouring gate according to the filling rate balance state so as to generate the updated flow distribution scheme.
  8. 8. The intelligent monitoring method for casting production based on the internet of things according to claim 1, wherein the calculating internal stress distribution caused by casting filling integrity and stagnation area pressure fluctuation according to the updated flow distribution scheme to obtain an internal defect prediction result comprises: The method comprises the steps of updating a flow distribution scheme of casting liquid, constructing a three-dimensional finite element model of the casting according to the updated flow distribution scheme, calculating the flow speed and pressure distribution of the casting liquid according to the three-dimensional finite element model of the casting liquid to generate filling rate distribution data, calculating a comprehensive stress value according to the filling rate distribution data and the temperature gradient of a stagnant area, and generating an internal defect prediction result comprising the bubble generation probability and the bubble size distribution range according to the comprehensive stress value and the temperature change rate of the solidification process of the casting liquid.
  9. 9. The intelligent monitoring method for casting production based on the internet of things according to claim 8, wherein the generating the internal defect prediction result including the bubble generation probability and the bubble size distribution range according to the integrated stress value and the temperature change rate of the casting solution solidification process comprises: and estimating a bubble diameter range according to the difference value between the comprehensive stress value and the critical nucleation stress threshold value and the temperature change rate in the solidification process of the casting solution, and generating the bubble size distribution range.
  10. 10. The intelligent monitoring method for casting production based on the internet of things according to claim 1, wherein if the bubble size distribution range in the internal defect prediction result exceeds a preset threshold, iteratively adjusting the gate opening time sequence deviation and the flow distribution ratio to obtain the optimized casting parameters, comprising: If the maximum value of the bubble size distribution range in the internal defect prediction result exceeds a preset size threshold, extracting the current gate opening time sequence deviation value and flow distribution proportion data, and adjusting the opening time and flow distribution proportion of the corresponding gate according to the position information of the bubble size exceeding area to obtain the optimized casting parameters.

Description

Casting production intelligent monitoring method based on Internet of things Technical Field The invention relates to the technical field of information, in particular to an intelligent casting production monitoring method based on the Internet of things. Background The casting production is the core field of manufacturing industry, and is directly related to quality and efficiency of industries such as machinery, aviation, automobiles and the like, and the production precision and stability are greatly improved by introducing intelligent monitoring technology. However, when the existing method is used for controlling the flow of casting liquid under a complex casting structure, the existing method is difficult to adapt to dynamically-changed process conditions, so that the quality of the casting is unstable. Particularly, when the geometric characteristics of the casting mould or the viscosity of casting liquid change, the traditional monitoring system is difficult to accurately analyze and adjust the technological parameters in real time, and the casting forming effect is often influenced due to response lag or rough regulation. In V-method casting, flow distribution of a multi-gate system is a key technical link. The change of the geometric complexity of the casting mold can directly influence the flow path and the speed distribution of the casting liquid, and if the flow ratio and the opening time sequence of each pouring gate cannot be dynamically adjusted, the partial area can be unevenly filled. For example, in complex moulds, certain areas may be subject to bubbles due to stagnant casting liquid, which in turn may cause internal defects and even stress concentrations due to imbalances in filling rate. This imbalance in flow distribution is not only due to the complexity of the mold structure, but is also closely related to the dynamic changes in the flow characteristics of the casting solution. Fluctuations in flow characteristics exacerbate the non-linear variation in flow distribution, disabling conventional fixed distribution strategies and making uniform filling and defect suppression difficult. Therefore, how to analyze the geometric complexity of the casting mold and the flow characteristics of the casting liquid in real time through the technology of the Internet of things dynamically adjusts the flow distribution proportion and the opening time sequence of the multi-gate system so as to balance the filling rate and reduce the bubble generation and the stress concentration, thereby becoming a key problem for improving the quality of the casting. Disclosure of Invention The invention provides an intelligent casting production monitoring method based on the Internet of things, which mainly comprises the following steps: Preprocessing casting mold geometric data and casting liquid viscosity distribution data, determining a temperature gradient of a stagnation area and a local difference of casting liquid viscosity according to extracted casting liquid flow state parameters, and analyzing geometric constraints of the stagnation area of a casting mold structure to obtain a local gate pressure gradient of each gate area; If the local difference of the viscosity of the casting solution exceeds the preset bubble generation critical viscosity, determining the bubble generation probability and the bubble size distribution range, and if the local difference of the viscosity of the casting solution does not exceed the preset bubble, generating the critical viscosity, maintaining the current gate opening time sequence and flow distribution; Optimizing a multi-gate opening time sequence according to the bubble generation probability and the bubble size distribution range, determining the opening sequence of each gate and the gate opening time sequence deviation, and generating an optimized opening time sequence scheme; dynamically adjusting the flow distribution proportion of each gate according to the optimized opening time sequence scheme to generate a flow distribution initial scheme; Acquiring casting liquid flow state feedback data after executing the initial flow distribution scheme, and generating an updated flow distribution scheme; Calculating the filling integrity of the casting and the internal stress distribution caused by the pressure fluctuation of the stagnation area according to the updated flow distribution scheme to obtain an internal defect prediction result; And if the bubble size distribution range in the internal defect prediction result exceeds a preset threshold, iteratively adjusting the gate opening time sequence deviation and the flow distribution proportion to obtain optimized casting parameters, and if the bubble size distribution range does not exceed the preset threshold, confirming the casting parameters and outputting a final casting scheme. Further, the preprocessing of the geometric data of the casting mold and the viscosity distribution data of the casting liquid, determ