Method and system for predicting sealing life of lithium iron phosphate rotary kiln

Abstract

The invention belongs to the technical field of industrial equipment monitoring and control, and particularly relates to a lithium iron phosphate rotary kiln sealing life prediction method and system, and the method comprises the following steps: S1, collecting multi-dimensional sensor data in the operation process of a rotary kiln, carrying out the angular domain resampling synchronization processing of the multi-dimensional sensor data, and obtaining a standardized angular domain gap waveform sequence; and S2, according to the multi-dimensional sensor data and the angular domain gap waveform sequence, constructing a thermogravity coupling collapse factor to characterize the degree of plastic deformation after stripping heat influence, and constructing a sealing impact wear index to characterize the damage strength of the current working condition to the sealing element. Through physical feature decoupling and nonlinear damage evaluation, the problem of sealing failure prediction lag caused by cylinder deformation under the high-temperature working condition is effectively solved, and safe and conservative life prediction and active life extension control are achieved.

Description

Technical Field

[0001] This invention belongs to the field of industrial equipment monitoring and control technology, specifically relating to a method and system for predicting the seal life of a lithium iron phosphate rotary kiln. Background Technology

[0002] As a mainstream cathode material for new energy batteries, lithium iron phosphate's high-temperature calcination process is crucial to product performance. The rotary kiln, as the core equipment in this process, directly determines the electrochemical performance of the product through the stability of its internal atmosphere. To prevent air ingress or process gas leakage, mainstream equipment manufacturers typically employ a composite sealing structure combining mechanical and gas seals at the kiln head and tail to ensure a sealed production environment.

[0003] However, in actual industrial production, rotary kilns face extreme operating conditions. The kiln shell is typically supported by bottom rollers at intervals, resulting in a long span of the shell being suspended between the two supports. Under the combined effect of prolonged high-temperature electric heating and its own gravity, the suspended section of the shell undergoes irreversible metal creep, producing a slight downward collapse deformation. This deformation causes the sealing interface to evolve from an ideal concentric circle trajectory to a complex eccentric elliptical trajectory during rotation, resulting in severe periodic radial runout and exacerbating the wear of the seals.

[0004] Existing seal life prediction technologies typically rely solely on operating time or a single temperature threshold, exhibiting the following logical flaws: First, current technologies cannot distinguish between temporary thermal expansion of the cylinder due to heating and permanent plastic collapse caused by gravity, leading to misjudgments of seal gap changes—that is, they cannot accurately isolate the interference of thermoelastic deformation. Second, existing technologies often assume linear wear, neglecting the fact that when cylinder deformation causes seal gap fluctuations, the impact force on the sealing ring increases exponentially with rotational speed and deformation, causing the prediction model to lag significantly in the later stages of failure. Therefore, a method capable of decoupling physical deformation characteristics and accurately predicting life is urgently needed. Summary of the Invention

[0005] This invention provides a method and system for predicting the seal life of a lithium iron phosphate rotary kiln, in order to solve the technical problems of existing technologies being unable to isolate thermoelastic deformation interference and lacking nonlinear damage assessment, which leads to prediction lag and misjudgment.

[0006] In a first aspect, the present invention provides a method for predicting the sealing life of a lithium iron phosphate rotary kiln, comprising the following steps: S1: Collect multi-dimensional sensor data during the operation of the rotary kiln, perform corner domain resampling and synchronous processing on the multi-dimensional sensor data, and obtain a standardized corner domain gap waveform sequence. S2, based on multi-dimensional sensor data and angular domain gap waveform sequence, a thermo-gravity coupling collapse factor is constructed to characterize the degree of plastic deformation after peeling off the heat effect, and a sealing impact wear index is constructed to characterize the damage intensity of the seal under the current working condition. S3, the thermogravity coupling collapse factor, sealing impact wear index and the rate of change of gas seal supply pressure are used to form a time series vector, which is then input into a pre-trained long short-term memory network model to output the predicted remaining seal life. S4 compares the predicted remaining seal life with the maintenance warning threshold. If it is lower than the maintenance warning threshold, it generates control commands that include reducing the rotation speed and increasing the gas seal supply pressure to adaptively control the rotary kiln.

[0007] Furthermore, the multi-dimensional sensor data includes sealing gap data, cylinder surface temperature, real-time rotational speed of the rotary kiln, and gas seal supply pressure; Corner-domain resampling synchronization processing of multidimensional sensor data includes: By utilizing the real-time rotation speed of the rotary kiln, the collected sealing gap data is mapped onto the rotation angle of the rotary kiln. The rotary kiln is divided into N equal angles for each rotation. The sealing gap value corresponding to each angle is obtained by interpolation algorithm, forming an angle domain gap waveform sequence.

[0008] Furthermore, the thermogravity coupling collapse factor satisfies the following relationship:

[0009] In the formula, Let be the thermogravimetric coupling collapse factor at time t. Let be the effective value of the radial runout of the sealing gap at time t. Let be the surface temperature of the cylinder at time t. The standard operating temperature for rotary kilns. This is the thermal compensation coefficient. It is an exponential function with the natural constant as its base.

[0010] Furthermore, the effective value of radial runout The method of obtaining it is: Calculate the standard deviation of the angular domain gap waveform sequence at N sampling points within one revolution of the rotary kiln, and use the calculated standard deviation as the effective value of radial runout.

[0011] Furthermore, the seal impact wear index satisfies the following relationship:

[0012] In the formula, Let be the seal impact wear index at time t. This refers to the real-time rotational speed of the rotary kiln. The current gas supply pressure for the gas seal. Set the target pressure for the gas seal. To prevent the protection constant from being zero, The thermogravity coupling collapse factor at time t.

[0013] Furthermore, the Long Short-Term Memory (LSTM) network model is trained using asymmetric weighted mean squared error as the loss function, which satisfies the following relationship:

[0014] In the formula, This is the total loss value. The total number of samples, The remaining lifetime predicted by the model. For the actual remaining lifespan, To overestimate the risk penalty weight, This is to underestimate the risk penalty weight.

[0015] Furthermore, overestimating the risk penalty weights is greater than underestimating them, causing the prediction results of the Long Short-Term Memory Network model to tend towards conservatism and safety.

[0016] Furthermore, the control commands to reduce the rotational speed and increase the gas supply pressure to the gas seal are used to reduce the value of the seal impact wear index, thereby slowing down the rate of decay of the seal's remaining life.

[0017] Furthermore, the standard operating temperature of the rotary kiln is a constant, the thermal compensation coefficient is set to a positive value based on the thermal expansion properties of the cylinder material, and the zero-prevention protection constant is a non-zero positive number to prevent the denominator from being zero.

[0018] Secondly, the present invention provides a lithium iron phosphate rotary kiln seal life prediction system, including a memory and a processor. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned lithium iron phosphate rotary kiln seal life prediction method is implemented.

[0019] The beneficial effects are as follows: By constructing a thermo-gravity coupled collapse factor, this invention effectively eliminates the interference of thermal expansion and enables early identification of minute plastic deformations in the cylinder; by constructing a sealing impact wear index, it assesses the nonlinear influence of rotational speed and air pressure on sealing damage, thereby improving prediction accuracy; by using an asymmetric loss function to train the model, it ensures the safety and conservatism of the prediction results; and by combining an adaptive control strategy, it can proactively delay the occurrence of faults after discovering potential hazards, maximizing the continuous operation benefits of the production line. Attached Figure Description

[0020] Figure 1 This is a flowchart of the present invention.

[0021] Figure 2 This is a diagram illustrating the effect of the wear index monitoring and adaptive control process in this invention.

[0022] Figure 3 This is a comparison chart showing the accuracy of the seal life prediction between the present invention and existing technologies. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] An embodiment of the method for predicting the seal life of a lithium iron phosphate rotary kiln provided by the present invention: like Figure 1 As shown, the method for predicting the seal life of a lithium iron phosphate rotary kiln includes the following steps: S1 collects multi-dimensional sensor data during the operation of the rotary kiln, performs corner domain resampling and synchronous processing on the multi-dimensional sensor data, and obtains a standardized corner domain gap waveform sequence.

[0025] Specifically, four non-contact eddy current displacement sensors are evenly arranged circumferentially at the kiln head and kiln tail sealing devices of the rotary kiln to collect the dynamic gap value of the sealing interface in real time, denoted as . Simultaneously, the surface temperature of the cylinder is acquired through the equipment's PLC control system. Rotary kiln real-time rotation speed and the air seal supply pressure .

[0026] Because the rotary kiln rotation speed fluctuates with process requirements, data acquired in the time domain cannot be directly aligned. Therefore, angular domain resampling is necessary. Using the rotation speed signal, all acquired gap data are mapped to the rotation angle of the rotary kiln, ranging from 0 to 360 degrees. Each rotation of the rotary kiln is divided into N equal angles, for example, N=360. An interpolation algorithm is used to obtain the sealing gap value corresponding to each angle, forming a standardized angular domain gap waveform sequence. ,in Indicates the first A point of view This indicates the current time period.

[0027] For example, suppose the rotary kiln is currently rotating at 3 revolutions per minute, meaning it completes one revolution every 20 seconds. If the sensor sampling rate is 10Hz, then 200 points are collected per revolution. Using an interpolation algorithm, these are resampled to 360 points, corresponding to degrees 0 to 359. If the rotation speed changes to 6 revolutions per minute in the next moment, 100 points are collected per revolution, again resampled to 360 points. In this way, regardless of the rotation speed change, the data length for each revolution remains a fixed 360 points, achieving data alignment.

[0028] By resampling the collected data in the angular domain, the influence of rotational speed fluctuations on data alignment was eliminated, ensuring the consistency of subsequent feature extraction in the spatial dimension and providing a standardized data foundation for accurate analysis of cylinder deformation.

[0029] S2, based on multidimensional sensor data and angular domain gap waveform sequence, constructs a thermo-gravity coupling collapse factor to characterize the degree of plastic deformation after peeling off the thermal effect, and constructs a sealing impact wear index to characterize the damage intensity of the seal under the current working condition.

[0030] This step aims to extract the physical characteristics that truly affect the seal life from the raw data. The seal gap fluctuation of the rotary kiln includes recoverable thermal expansion and irreversible gravity collapse. Thermal expansion is elastic deformation, while gravity collapse is plastic deformation. Directly using the total gap fluctuation will introduce thermal expansion noise, so it is necessary to construct an index to eliminate the thermal effect.

[0031] Specifically, the thermogravity coupling collapse factor is first constructed. It satisfies the following relationship:

[0032] in, The effective value of the radial runout of the sealing gap at time t is obtained by calculating the standard deviation of N sampling points within one revolution; The standard operating temperature for rotary kilns. This is the thermal compensation coefficient.

[0033] For example, assuming the rotary kiln is designed to be at a certain temperature. thermal compensation coefficient .

[0034] Scenario A: The equipment has just been started, and the temperature is... The effective value of radial runout was measured. At this time, the exponent term ,but .

[0035] Scenario B: The equipment is operating at high temperature, and the temperature rises to... Due to thermal expansion, the measured effective value of radial runout increased to At this time, the exponent term ,but .

[0036] As can be seen, although the original runout value increased from 2mm to 4mm, the thermogravity coupling collapse factor after temperature compensation did not increase significantly, and even decreased slightly. This indicates that the increase in runout is mainly caused by thermal expansion, rather than a permanent collapse of the cylinder. If the temperature remains constant while the effective value of radial runout increases, the thermogravity coupling collapse factor will increase significantly, thus accurately identifying plastic deformation.

[0037] Furthermore, a sealing impact wear index was constructed. The wear rate of the mechanical seal increases nonlinearly after the cylinder collapses, and is directly proportional to the square of the rotational speed and inversely proportional to the film pressure; this index satisfies the following relationship:

[0038] in, Set the target pressure for the gas seal. To prevent the protection constant from being zero, for example, a value of 0.1 is used.

[0039] For example, using the calculations above ,set up , .

[0040] Operating Condition 1: Real-time rotational speed of the rotary kiln rpm, normal gas seal supply pressure ; .

[0041] Operating Condition 2: Speed ​​increased to rpm, air pressure drops to ; .

[0042] It can be seen that the deterioration of the working conditions caused the wear index to surge from 11.94 to 138.24. This index accurately assesses the devastating wear under high-speed, low-pressure, and large-deformation working conditions.

[0043] By constructing a thermogravity coupled collapse factor and a sealing impact wear index, the physical characteristics were decoupled and evaluated from the two dimensions of deformation mechanism and damage mechanism, respectively. This enables the early detection of hidden dangers and the keen capture of the nonlinear impact of operating condition changes on lifespan, significantly improving the effectiveness of the characteristics.

[0044] S3 constructs a time series vector from the thermogravity coupling collapse factor, the sealing impact wear index, and the rate of change of the gas seal supply pressure, and inputs it into a pre-trained long short-term memory network model to output the predicted remaining seal life.

[0045] Specifically, a two-layer long short-term memory network is constructed, with the input layer receiving time series vectors. ,in The output layer outputs the predicted remaining seal life, representing the rate of change in air pressure. .

[0046] To ensure the safety of industrial production, asymmetric weighted mean square error is used as the loss function to train the model, as shown in the formula:

[0047] set up , Assuming actual lifespan Hour.

[0048] Scenario A (Overestimation, Danger): Model Prediction Hour; .

[0049] Scenario B (underestimation, safe): Model prediction Hour; .

[0050] As can be seen, for the same prediction error of 20 hours, the penalty for overestimating lifespan is 10 times that for underestimating it. This mechanism forces the model to tend to give slightly smaller predictions than the true values ​​during training, thus leaving a safety margin.

[0051] like Figure 3 As shown in the figure, the present invention is compared with the prior art. The prior art, due to the failure to eliminate thermal expansion, has a prediction curve that fluctuates wildly and lags behind; while the prediction curve of the present invention is smooth and always closely follows the actual remaining life of the equipment, verifying the accuracy and conservative safety characteristics of the model.

[0052] By introducing a physically enhanced LSTM model and combining it with an asymmetric loss function, not only is the accuracy of the prediction improved, but the prediction results are also ensured to comply with the principle of industrial safety: it is better to be conservative than reckless, thus effectively avoiding unexpected downtime accidents.

[0053] S4 compares the predicted remaining seal life with the maintenance warning threshold. If it is lower than the maintenance warning threshold, it generates control commands that include reducing the rotation speed and increasing the gas seal supply pressure to adaptively control the rotary kiln.

[0054] Specifically, the system monitors the remaining lifespan of the seal in real time. ,when When the speed drops below the preset maintenance warning threshold, it indicates that the seal is about to fail. At this time, the system does not immediately shut down, but instead sends a command to the PLC controller to gradually reduce the real-time rotation speed of the rotary kiln. And appropriately increase the gas supply pressure of the gas seal. .

[0055] Reference Figure 2 When the seal impact wear index reaches the warning threshold, the system triggers a control strategy, causing the seal impact wear index to drop rapidly and significantly and remain at a low level. This is because, according to the formula for the seal impact wear index, both reducing the rotational speed and increasing the air pressure can significantly reduce the wear index. This allows... Figure 3 The slope of the actual remaining life curve of the equipment becomes gentler after the control point, successfully extending the equipment's operating time until the planned maintenance window is reached.

[0056] By combining life prediction results with adaptive control strategies, a shift from passive monitoring to proactive life extension is achieved, enabling equipment to operate even with defects without interrupting production, thus maximizing the economic benefits and production continuity of the equipment.

[0057] An embodiment of the lithium iron phosphate rotary kiln seal life prediction system provided by the present invention: The lithium iron phosphate rotary kiln seal life prediction system includes a processor and a memory. The memory stores computer program instructions, which are executed by the processor to implement the above-mentioned lithium iron phosphate rotary kiln seal life prediction method.

[0058] The lithium iron phosphate rotary kiln seal life prediction system also includes other components well known to those skilled in the art, such as communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

[0059] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions stored or otherwise maintained by such a computer-readable medium.

[0060] The above are all preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Therefore, all equivalent changes made in accordance with the structure, shape and principle of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for predicting the seal life of a lithium iron phosphate rotary kiln, characterized in that, Includes the following steps: S1: Collect multi-dimensional sensor data during the operation of the rotary kiln, perform corner domain resampling and synchronous processing on the multi-dimensional sensor data, and obtain a standardized corner domain gap waveform sequence. S2, based on multi-dimensional sensor data and angular domain gap waveform sequence, a thermo-gravity coupling collapse factor is constructed to characterize the degree of plastic deformation after peeling off the heat effect, and a sealing impact wear index is constructed to characterize the damage intensity of the seal under the current working condition. S3, the thermogravity coupling collapse factor, sealing impact wear index and the rate of change of gas seal supply pressure are used to form a time series vector, which is then input into a pre-trained long short-term memory network model to output the predicted remaining life of the seal. S4 compares the predicted remaining seal life with the maintenance warning threshold. If it is lower than the maintenance warning threshold, it generates control commands that include reducing the rotation speed and increasing the gas seal supply pressure to adaptively control the rotary kiln.

2. The method for predicting the seal life of a lithium iron phosphate rotary kiln according to claim 1, characterized in that, Multidimensional sensor data includes sealing gap data, cylinder surface temperature, real-time rotational speed of the rotary kiln, and gas seal supply pressure; Corner-domain resampling synchronization processing of multidimensional sensor data includes: By utilizing the real-time rotation speed of the rotary kiln, the collected sealing gap data is mapped onto the rotation angle of the rotary kiln. The rotary kiln is divided into N equal angles for each rotation. The sealing gap value corresponding to each angle is obtained by interpolation algorithm, forming an angle domain gap waveform sequence.

3. The method for predicting the seal life of a lithium iron phosphate rotary kiln according to claim 1, characterized in that, The thermogravity coupling collapse factor satisfies the following relationship: In the formula, Let be the thermogravimetric coupling collapse factor at time t. Let be the effective value of the radial runout of the sealing gap at time t. Let be the surface temperature of the cylinder at time t. The standard operating temperature for rotary kilns. This is the thermal compensation coefficient. It is an exponential function with the natural constant as its base.

4. The method for predicting the seal life of a lithium iron phosphate rotary kiln according to claim 3, characterized in that, Effective value of radial runout The method of obtaining it is: Calculate the standard deviation of the angular domain gap waveform sequence at N sampling points within one revolution of the rotary kiln, and use the calculated standard deviation as the effective value of radial runout.

5. The method for predicting the seal life of a lithium iron phosphate rotary kiln according to claim 3, characterized in that, The seal impact wear index satisfies the following relationship: In the formula, Let be the seal impact wear index at time t. This refers to the real-time rotational speed of the rotary kiln. The current gas supply pressure for the gas seal. Set the target pressure for the gas seal. To prevent the protection constant from being reduced to zero, The thermogravity coupling collapse factor at time t.

6. The method for predicting the seal life of a lithium iron phosphate rotary kiln according to claim 1, characterized in that, The Long Short-Term Memory (LSTM) network model is trained using asymmetric weighted mean squared error as the loss function, which satisfies the following relationship: In the formula, This is the total loss value. The total number of samples, The remaining lifetime predicted by the model. For the actual remaining lifespan, To overestimate the risk penalty weight, This is to underestimate the risk penalty weight.

7. The method for predicting the seal life of a lithium iron phosphate rotary kiln according to claim 6, characterized in that, Overestimating the risk penalty weights is greater than underestimating them, causing the prediction results of the Long Short-Term Memory Network model to tend to be conservative and safe.

8. The method for predicting the seal life of a lithium iron phosphate rotary kiln according to claim 1, characterized in that, The control commands to reduce the rotational speed and increase the gas supply pressure to the gas seal are used to reduce the value of the seal impact wear index, thereby slowing down the rate of decay of the seal's remaining life.

9. The method for predicting the seal life of a lithium iron phosphate rotary kiln according to claim 5, characterized in that, The standard operating temperature of the rotary kiln is a constant. The thermal compensation coefficient is set to a positive value based on the thermal expansion properties of the cylinder material. The zero protection constant is a non-zero positive number to prevent the denominator from being zero.

10. A lithium iron phosphate rotary kiln seal life prediction system, characterized in that, It includes a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the method for predicting the sealing life of a lithium iron phosphate rotary kiln as described in any one of claims 1-9 is implemented.

Images