Casting mold ingot stripping detection method and system

Abstract

The invention relates to the technical field of cast ingot demoulding of a casting machine, in particular to a method and system for detecting demoulding of a casting mould, and the method comprises the following steps: constructing a voiceprint feature database according to voiceprint features corresponding to resonance echo signals generated under different demoulding degrees after the casting mould is knocked once; sound wave signals sent after the target casting mold is knocked once are collected in real time; separating the sound wave signal to obtain a resonance echo signal of the target casting mold, extracting voiceprint features of the separated resonance echo signal, and recognizing the ingot stripping degree of the target casting mold according to a voiceprint feature database; and judging whether the target casting mold is stripped or not based on the stripping degree. According to the invention, the ingot stripping degree is identified by using the resonance echo signal generated after the casting mould is knocked once, and whether the casting mould is stripped or not is judged according to the ingot stripping degree, so that the frequency of impact on the casting mould can be reduced, and low-loss detection is realized; due to the fact that detection is carried out during knocking demolding, the situation that the aluminum ingots fall off and damage the detection device in the returning process can be avoided, and the risk is lower.

Description

Technical Field

[0001] The present invention relates to the technical field of ingot demoulding of a casting machine, and in particular to a method and system for detecting ingot demoulding of a casting mold. Background Art

[0002] In the aluminum alloy smelting industry, the aluminum alloy solution smelted in the furnace is poured into the mold through a distributor to form an aluminum alloy ingot. After cooling, it is knocked out of the mold by a demoulding device and falls onto the ingot conveyor chain and finally transported to the stacking.

[0003] In the demolding operation of aluminum ingots, the mold is mostly knocked once at present, but there are often aluminum ingots that cannot be completely demolded, which increases the workload of the staff. If not handled in time, the undemolded aluminum ingots may fall to the bottom of the continuous casting machine during the return process, jam the continuous casting machine, and make the continuous casting machine unable to operate normally or even damage the continuous casting machine. If the aluminum ingot does not fall off the mold during the return process, then when the mold returns to the distributor position, since there is already an aluminum ingot in the mold, the repeated pouring of new aluminum alloy solution will cause a large amount of aluminum alloy solution to overflow, posing certain safety hazards.

[0004] The "Study on Diagnosis of Demolding Failure of Aluminum Ingots Based on Acoustic Signals" provided by Li Jianhua et al. discloses a diagnostic scheme for demolding failure of ingots based on acoustic signals. This diagnostic scheme intermittently taps the same mold, obtains the sound signals of the first and second taps, processes them, calculates the peak frequency difference of the two sound signals, and compares the peak frequency difference with the set threshold to diagnose whether the aluminum ingot is demolded before the second tap. The implementation of this scheme depends on tapping each mold twice, which is bound to accelerate the wear of the mold and may affect the final quality of the aluminum ingot. Moreover, since the mold is usually made of gray cast iron, although it has excellent shock absorption, it has low strength, so repeated impacts are prone to cracks, greatly reducing its service life.

[0005] The Chinese invention patent with publication number CN116618591B provides a demolding control method for aluminum ingots that have not been demolded by a casting machine. This method is similar to the non-demolding detection device disclosed in "Design and Installation of Non-Demolding Detection Alarm and Secondary Strike System for 20kg Aluminum Ingot Continuous Casting Machine" provided by Zhang Xiuru et al., and a non-demolding detection device for ingots provided by the Chinese utility model with publication number CN20506819U. Both measure the depth of the mold during the return process, but the difference lies in the different measurement methods. Among them, "Research on Demolding Fault Diagnosis of Aluminum Ingots Based on Acoustic Signals" measures the depth of the mold during the return process using a rangefinder to determine whether there are still aluminum ingots that have not been demolded in the mold. However, since the aluminum ingots that have not been demolded may fall at any time during the return process, if the rangefinder lacks protection, it may be damaged by the falling aluminum ingots. Similarly, the measurement methods provided by "Design and Installation of Non-Demolding Detection Alarm and Secondary Strike System for 20kg Aluminum Ingot Continuous Casting Machine" and a non-demolding detection device for ingots have this problem.

[0006] Based on this, it is necessary to provide a low-loss, low-risk method for detecting mold stripping. Summary of the invention

[0007] The object of the present invention is to provide a method and system for detecting mold demolding, which utilizes the resonant echo signal generated after the mold is knocked once to identify the degree of demolding, and then determines whether the mold is demolded based on the degree of demolding. This can reduce the number of impacts on the mold and achieve low-loss detection. In addition, since the detection is performed when the mold is demolded by knocking, it can also avoid the aluminum ingots from falling and damaging the detection device during the return process, which has a lower risk, thereby solving the problems pointed out in the background technology.

[0008] The embodiment of the present invention is achieved through the following technical solution: A casting mold stripping detection method comprises the following steps:

[0009] A voiceprint feature database is constructed based on the voiceprint features corresponding to the resonant echo signals generated at different degrees of stripping after the mold is struck once;

[0010] Real-time collection of the sound wave signal emitted by the target casting mold after it is struck once;

[0011] Separating the acoustic wave signal to obtain a resonant echo signal of a target casting mold, extracting a voiceprint feature of the separated resonant echo signal, and identifying a degree of ingot stripping of the target casting mold according to the voiceprint feature database;

[0012] Whether the target mold is debonded is determined based on the debonding degree.

[0013] According to a preferred embodiment, determining whether the target casting mold is debonded based on the debonding degree specifically includes:

[0014] It is determined whether the degree of stripping is less than a preset stripping threshold value. If so, it is determined that the target mold is not stripped; otherwise, it is determined that the target mold is stripped.

[0015] According to a preferred embodiment, after determining whether the target mold is debonded based on the debonding degree, the method further includes verifying whether the target mold is debonded, and updating the debonding state of the target mold based on the verification result, wherein the verification of whether the target mold is debonded is specifically as follows:

[0016] Obtaining the time interval between adjacent ingots being demoulded from the ingot conveyor chain within a set time range after the target mold is struck once;

[0017] Determine whether the time interval is greater than a preset time threshold. If so, determine that the target mold has not been stripped, and generate a secondary stripping instruction, which is used to instruct the secondary stripping device to perform a secondary knocking action on the target mold. Otherwise, determine that the target mold has been stripped.

[0018] According to a preferred embodiment, after real-time acquisition of the sound wave signal emitted by the target casting mold after being struck once, the method further includes performing noise reduction processing on the sound wave signal.

[0019] According to a preferred embodiment, the method further comprises enhancing the sound wave signal after the noise reduction process.

[0020] The present invention also provides a casting mold stripping detection system, comprising:

[0021] The database construction module is configured to construct a voiceprint feature database according to the voiceprint features corresponding to the resonant echo signals generated at different degrees of stripping after the mold is struck once;

[0022] A data acquisition module is configured to collect in real time a sound wave signal emitted by the target casting mold after being struck once;

[0023] a signal separation module, configured to separate the acoustic wave signal to obtain a resonant echo signal of a target casting mold, extract a voiceprint feature of the separated resonant echo signal, and identify a degree of debonding of the target casting mold according to the voiceprint feature database;

[0024] The pattern matching module is configured to determine whether the target casting mold is debonded based on the debonding degree.

[0025] According to a preferred embodiment, the pattern matching module is configured to determine whether the degree of stripping is less than a preset stripping threshold, and if so, determine that the target mold is not stripped, otherwise, determine that the target mold is stripped.

[0026] According to a preferred embodiment, the system further includes a verification module, which is configured to verify whether the target casting mold is debonded, and update the debonded state of the target casting mold based on the verification result. The verification of whether the target casting mold is debonded is specifically as follows:

[0027] Obtaining the time interval between adjacent ingots being demoulded from the ingot conveyor chain within a set time range after the target mold is struck once;

[0028] Determine whether the time interval is greater than a preset time threshold. If so, determine that the target mold has not been stripped, and generate a secondary stripping instruction, which is used to instruct the secondary stripping device to perform a secondary knocking action on the target mold. Otherwise, determine that the target mold has been stripped.

[0029] According to a preferred embodiment, the system further comprises a noise reduction module, wherein the noise reduction module is configured to perform noise reduction processing on the sound wave signal after real-time acquisition of the sound wave signal emitted by the target casting mold after being struck once.

[0030] According to a preferred embodiment, the system further comprises a signal enhancement module, and the signal enhancement module is configured to enhance the sound wave signal after the noise reduction process.

[0031] The technical solution of a mold debonding detection method and system provided by the present invention has at least the following advantages and beneficial effects: (1) The degree of debonding is identified by utilizing the resonant echo signal generated after the mold is struck once, and then whether the mold is demolded is determined based on the degree of debonding, thereby reducing the number of impacts on the mold and achieving low-loss detection; (2) Since the detection is performed when the mold is debonded, the aluminum ingots can be prevented from falling and damaging the detection device during the return process, which reduces the risk. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A schematic flow chart of a casting mold stripping detection method provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0034] Example 1

[0035] Figure 1 The following is a flow chart of a casting mold stripping detection method provided by an embodiment of the present invention. Figure 1As shown, the casting mold stripping detection method includes steps S1 to S4, and steps S1 to S4 are further described below:

[0036] Step S1, constructing a voiceprint feature database;

[0037] Step S11, obtaining the resonant echo signal generated at different degrees of stripping after the mold is struck once.

[0038] In this embodiment, different degrees of ingot stripping are defined according to the size of the separation gap between the mold and the ingot in the trough, and can be divided into multiple levels such as complete separation, incomplete separation and no separation at all. Complete separation means that the separation gap extends from the top of the trough to the bottom of the trough, completely penetrating the connecting surface between the mold and the ingot, and the ingot can be detached from the mold by itself after the mold is flipped; incomplete separation means that the separation gap does not completely penetrate the connecting surface between the mold and the ingot, and the ingot may be detached from the mold after the mold is flipped due to gravity, demolding conveyor chain or other external environmental vibrations; no separation at all means that there is no gap at the connecting surface between the mold and the ingot, and the ingot cannot be detached from the mold after the mold is flipped.

[0039] When the ingot is completely separated, since there is a completely penetrating gap between the ingot in the trough and the mold, the mold will continue to vibrate at high frequency for a period of time, and the sound waves generated by the knocking will be reflected multiple times between the gaps, with a large attenuation, and the resonant echo signal will show very significant characteristics; when the ingot is not completely separated, the attenuation degree is lower than that when the ingot is completely separated, and there is a certain degree of difference between the resonant echo signal generated when the ingot is completely separated; when the ingot is not completely separated, the mold and the ingot form a whole, and due to the shock absorption of the ingot, the resonant echo signal will disappear instantly.

[0040] Step S12: construct a voiceprint feature database according to the voiceprint features corresponding to the resonant echo signals generated at different degrees of stripping after the mold is struck once.

[0041] It should be noted that voiceprint recognition is a biometric technology, which is usually used to extract and select acoustic features of the speaker's voiceprint that have strong separability and high stability. This embodiment uses voiceprint recognition to extract the voiceprint features of the resonant echo signal generated under different degrees of stripping, and different degrees of stripping can be distinguished according to the voiceprint features.

[0042] This embodiment uses the resonant echo signal generated after the mold is knocked once to identify the degree of ingot stripping, and then determines whether to demold based on the degree of ingot stripping. This can reduce the number of impacts on the mold and achieve low-loss detection. In addition, since the detection is performed when the mold is knocked to strip, it can also avoid the aluminum ingot from falling and damaging the detection device during the return process, which reduces the risk.

[0043] Step S2, data collection;

[0044] In this embodiment, the sound wave signal emitted by the target mold after being struck once is collected in real time. It should be noted that the sound wave signal is complex in composition. In addition to the sound wave signal and the resonant echo signal emitted by the target mold after being struck once, it may also include the resonant echo after the adjacent mold is struck, the running noise of the demoulding conveyor chain, the running noise of the demoulding device and other environmental noises, so it is necessary to post-process it.

[0045] Step S3, post-processing of the acoustic wave signal;

[0046] In this embodiment, post-processing includes successively performing noise reduction processing, enhancement processing and separation on the sound wave signal to obtain the resonant echo signal of the target casting mold; further, extracting the voiceprint features of the separated resonant echo signal, performing pattern matching according to the voiceprint feature database, and identifying the degree of stripping of the target casting mold.

[0047] Step S4, pattern matching;

[0048] In this embodiment, pattern matching includes:

[0049] Step S41, determining whether the target mold is debonded based on the debonding degree, specifically comprising: determining whether the debonding degree is less than a preset debonding threshold, if so, determining that the target mold is not debonded, otherwise, determining that the target mold is debonded.

[0050] Step S42, verify whether the target mold is debonded, and update the debonding status of the target mold based on the verification result. The verification of whether the target mold is debonded is specifically as follows: obtain the time interval between adjacent ingots demoulded from the ingot conveyor chain within a set time range after the target mold is struck once; determine whether the time interval is greater than a preset time threshold, and if so, determine that the target mold is not debonded, issue a non-debonding alarm and generate a secondary debonding instruction, which is used to instruct the secondary demoulding device to perform a secondary knocking action on the target mold, otherwise, determine that the target mold is debonded.

[0051] Example 2

[0052] Based on the technical solution provided in the embodiment, the embodiment of the present invention provides a mold debonding detection system, including a database construction module, a data acquisition module, a signal separation module and a pattern matching module; wherein the database construction module is configured to construct a voiceprint feature database according to the voiceprint features corresponding to the resonant echo signals generated at different debonding degrees after the mold is struck once; the data acquisition module is configured to collect in real time the sound wave signals emitted by the target mold after it is struck once; the signal separation module is configured to separate the sound wave signals to obtain the resonant echo signals of the target mold, extract the voiceprint features of the separated resonant echo signals, and identify the debonding degree of the target mold according to the voiceprint feature database; the pattern matching module is configured to determine whether the target mold is debonded based on the debonding degree.

[0053] Specifically, the pattern matching module is configured to determine whether the degree of stripping is less than a preset stripping threshold value, and if so, determine that the target mold is not stripped, otherwise, determine that the target mold is stripped.

[0054] In addition, the system also includes a verification module, which is configured to verify whether the target mold is debonded, and update the debonding status of the target mold based on the verification result. The verification of whether the target mold is debonded is specifically as follows: obtaining the time interval between adjacent ingots being demolded on the ingot conveyor chain within a set time range after the target mold is struck once; judging whether the time interval is greater than a preset time threshold, and if so, determining that the target mold is not debonded, and generating a secondary debonding instruction, wherein the secondary debonding instruction is used to instruct the secondary demolding device to perform a secondary knocking action on the target mold, otherwise, determining that the target mold is debonded.

[0055] The system also includes a noise reduction module and a signal enhancement module. The noise reduction module is configured to, after real-time acquisition of the sound wave signal emitted by the target mold after a knock, perform noise reduction processing on the sound wave signal, including filtering out the resonance echo of the adjacent molds, the operating noise of the demoulding conveyor chain, and the operating noise of the demoulding device. The signal enhancement module is configured to enhance the sound wave signal after the noise reduction processing.

[0056] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for detecting mold stripping, characterized in that: The following steps are involved: A voiceprint feature database is constructed based on the voiceprint features corresponding to the resonant echo signals generated at different degrees of stripping after the mold is struck once; Real-time collection of the sound wave signal emitted by the target casting mold after it is struck once; Separating the acoustic wave signal to obtain a resonant echo signal of a target casting mold, extracting a voiceprint feature of the separated resonant echo signal, and identifying a degree of ingot stripping of the target casting mold according to the voiceprint feature database; Whether the target mold is debonded is determined based on the debonding degree.

2. The casting mold stripping detection method according to claim 1, characterized in that: Determining whether the target casting mold is debonded based on the debonding degree specifically includes: It is determined whether the degree of stripping is less than a preset stripping threshold value. If so, it is determined that the target mold is not stripped; otherwise, it is determined that the target mold is stripped.

3. The casting mold stripping detection method according to claim 2, characterized in that: After determining whether the target mold is debonded based on the debonding degree, the method further includes verifying whether the target mold is debonded, and updating the debonding state of the target mold based on the verification result, wherein the verification of whether the target mold is debonded is specifically as follows: Obtaining the time interval between adjacent ingots being demoulded from the ingot conveyor chain within a set time range after the target mold is struck once; Determine whether the time interval is greater than a preset time threshold. If so, determine that the target mold has not been stripped, and generate a secondary stripping instruction, which is used to instruct the secondary stripping device to perform a secondary knocking action on the target mold. Otherwise, determine that the target mold has been stripped.

4. The casting mold stripping detection method according to any one of claims 1 to 3, characterized in that: After real-time acquisition of the sound wave signal emitted by the target casting mold after being struck once, the method further includes performing noise reduction processing on the sound wave signal.

5. The casting mold stripping detection method according to claim 4, characterized in that: The method also includes enhancing the sound wave signal after the noise reduction process.

6. A casting mold stripping detection system, characterized in that: include: The database construction module is configured to construct a voiceprint feature database according to the voiceprint features corresponding to the resonant echo signals generated at different degrees of stripping after the mold is struck once; A data acquisition module is configured to collect in real time a sound wave signal emitted by the target casting mold after being struck once; a signal separation module, configured to separate the acoustic wave signal to obtain a resonant echo signal of a target casting mold, extract a voiceprint feature of the separated resonant echo signal, and identify a degree of debonding of the target casting mold according to the voiceprint feature database; The pattern matching module is configured to determine whether the target casting mold is debonded based on the debonding degree.

7. The casting mold stripping detection system according to claim 6, characterized in that: The pattern matching module is configured to determine whether the degree of stripping is less than a preset stripping threshold, and if so, determine that the target mold is not stripped, otherwise, determine that the target mold is stripped.

8. The casting mold stripping detection system according to claim 7, characterized in that: The system further includes a verification module, which is configured to verify whether the target casting mold is debonded, and update the debonded state of the target casting mold based on the verification result. The verification of whether the target casting mold is debonded is specifically as follows: Obtaining the time interval between adjacent ingots being demoulded from the ingot conveyor chain within a set time range after the target mold is struck once; Determine whether the time interval is greater than a preset time threshold. If so, determine that the target mold has not been stripped, and generate a secondary stripping instruction, which is used to instruct the secondary stripping device to perform a secondary knocking action on the target mold. Otherwise, determine that the target mold has been stripped.

9. The casting mold stripping detection system according to any one of claims 6 to 8, characterized in that: The system also includes a noise reduction module, which is configured to perform noise reduction processing on the sound wave signal after real-time acquisition of the sound wave signal emitted by the target casting mold after being struck once.

10. The casting mold stripping detection system according to claim 9, characterized in that: The system also includes a signal enhancement module, which is configured to enhance the sound wave signal after the noise reduction process.

Images