CN-115564020-B - Method, system, equipment and medium for determining lance position change state of oxygen lance

CN115564020BCN 115564020 BCN115564020 BCN 115564020BCN-115564020-B

Abstract

The application provides a method, a system, equipment and a medium for determining the change state of the gun position of an oxygen gun, wherein the method for determining the change state of the gun position of the oxygen gun comprises the steps of obtaining historical smelting data of a converter to construct a training set, training a deep reinforcement learning neural network through the constructed training set, the method and the device have the advantages that the real-time smelting data of the converter are obtained, and the gun position of the oxygen gun is determined through the trained deep reinforcement learning neural network according to the real-time smelting data.

Inventors

ZHAO MINGMING
CHEN BO
LIU BAOJU
ZHANG SEN
HE CHUNLAI

Assignees

中冶赛迪技术研究中心有限公司

Dates

Publication Date: 20260512
Application Date: 20220914

Claims (5)

1. The method for determining the change state of the lance position of the oxygen lance is characterized by comprising the following steps of: acquiring historical smelting data of a converter to construct a training set, wherein the historical smelting data comprises smelting process parameter historical data, oxygen lance position change state historical information and historical furnace mouth combustion images; training the deep reinforcement learning neural network through the constructed training set; acquiring real-time smelting data of a converter; Determining the gun position of the oxygen gun through a trained deep reinforcement learning neural network according to the real-time smelting data; Building a training set, comprising: constructing a smelting process parameter feature vector according to the smelting process parameter historical data; Constructing an oxygen lance position change state vector according to the oxygen lance position change state history information; constructing a reward variable according to the historical furnace mouth combustion image; and constructing a training set according to the smelting process parameter feature vector, the oxygen lance position change state vector and the rewarding variable.
2. The method of claim 1, wherein constructing a bonus variable based on the historical burner firing images comprises: Determining smelting states and smelting finished product quality indexes according to the historical furnace mouth combustion images; And constructing a reward variable according to the smelting state and the smelting finished product quality index.
3. An oxygen lance position change state determination system, comprising: The first acquisition module is used for acquiring historical smelting data of the converter to construct a training set, wherein the historical smelting data comprises smelting process parameter historical data, oxygen lance position change state historical information and historical furnace mouth combustion images, and the construction of the training set comprises the steps of constructing a smelting process parameter feature vector according to the smelting process parameter historical data, constructing an oxygen lance position change state vector according to the oxygen lance position change state historical information, constructing a reward variable according to the historical furnace mouth combustion images, and constructing the training set according to the smelting process parameter feature vector, the oxygen lance position change state vector and the reward variable; the training module trains the deep reinforcement learning neural network through the constructed training set; the second acquisition module is used for acquiring real-time smelting data of the converter; And the oxygen lance position determining module is used for determining the oxygen lance position through the trained deep reinforcement learning neural network according to the real-time smelting data.
4. An electronic device comprising one or more processors and storage means for storing one or more programs that, when executed by the one or more processors, cause the electronic device to implement the method of claim 1 or 2.
5. A computer readable storage medium, having stored thereon a computer program which, when executed by a processor of a computer, causes the computer to perform the method of claim 1 or 2.

Description

Method, system, equipment and medium for determining lance position change state of oxygen lance Technical Field The application relates to the technical field of steelmaking, in particular to a method, a system, equipment and a medium for determining a change state of a lance position of an oxygen lance. Background In the converter smelting process, the control of the lance position of the oxygen lance influences smelting quality indexes such as slag formation, decarburization, temperature rise and the like, and meanwhile, the control of the lance position of the oxygen lance is directly influenced with unstable production phenomena such as molten steel splashing, slag back drying and the like in the smelting process, so that the reasonable control of the lance position of the oxygen lance is very important in the converter steelmaking process. In general, in the whole converter smelting process, the lance position of the oxygen lance is controlled by manual operation of workers, so that the smelting quality is greatly dependent on the operation level and experience of operators, which seriously hinders the optimization and upgrading of the technical indexes of the steelmaking process. In particular, for some high-quality special steel grades, the requirements on the technological parameters are relatively high, the smelting quality needs to be strictly controlled, and even operators with abundant experience can have the problems of improper operation or misjudgment and the like, so that the smelting quality does not meet the standard, and serious economic loss is brought to a steel mill. In addition, the variability of manual operation often leads to unstable molten steel components smelted by different shifts or operators, which also brings some challenges to quality control of products. Disclosure of Invention In view of the above-mentioned drawbacks of the prior art, the present invention provides a method, system, device and medium for determining a lance position change state of an oxygen lance, so as to improve the stability and safety of production. The invention provides a method for determining a change state of a gun position of an oxygen gun, which comprises the following steps: acquiring historical smelting data of a converter to construct a training set, wherein the historical smelting data comprises smelting process parameter historical data, oxygen lance position change state historical information and historical furnace mouth combustion images; training the deep reinforcement learning neural network through the constructed training set; acquiring real-time smelting data of a converter; And determining the gun position of the oxygen gun through the trained deep reinforcement learning neural network according to the real-time smelting data. In an exemplary embodiment of the present application, constructing a training set includes: constructing a smelting process parameter feature vector according to the smelting process parameter historical data; Constructing an oxygen lance position change state vector according to the oxygen lance position change state history information; constructing a reward variable according to the historical furnace mouth combustion image; and constructing a training set according to the smelting process parameter feature vector, the oxygen lance position change state vector and the rewarding variable. In an exemplary embodiment of the application, constructing a bonus variable from the historical furnace mouth combustion image includes: Determining smelting states and smelting finished product quality indexes according to the historical furnace mouth combustion images; And constructing a reward variable according to the smelting state and the smelting finished product quality index. In a second aspect, the present application provides a method for controlling a lance position of an oxygen lance, comprising: Acquiring current smelting process parameters; obtaining the gun position change state of the oxygen gun according to the current smelting process parameters and a pre-trained deep reinforcement learning neural network, wherein the deep reinforcement learning neural network takes the smelting process parameters as input and takes the gun position change state of the oxygen gun as output for training; and changing the lance position of the oxygen lance according to the lance position changing state of the oxygen lance. In an exemplary embodiment of the present application, after changing the lance position according to the lance position changing state, the method further includes: Acquiring a real-time smelting finished product quality index and a real-time furnace mouth combustion image in a smelting process, wherein the real-time smelting finished product quality index and the real-time furnace mouth combustion image are obtained by smelting according to the current smelting process parameters based on the changed oxygen lance position; And taking the current