A lithium iron phosphate sintering furnace detection method, system, device and medium

Abstract

This invention discloses a method, system, equipment, and medium for detecting lithium iron phosphate sintering furnace pressure. The method includes: acquiring pressure data from historical sintering processes to construct at least one pressure range; collecting pressure data from at least one isothermal state during sintering; comparing the pressure data with the pressure range; if an anomaly is found, it indicates an abnormality in the sintering material or a problem with the sintering process; if no anomaly is found, it indicates that the sintering material is normal. Different batches of lithium iron phosphate powder have different sources, and their corresponding processes and raw materials are not entirely the same, resulting in lithium iron phosphate produced according to the original production process curve failing to meet requirements. This invention monitors pressure parameters during the sintering process and compares these parameters with historical sintering pressure data. If the pressure is higher or lower than the pressure range, it proves that the material is abnormal. This invention determines the material's qualification based on pressure parameters during the sintering process, saving manpower and resources for subsequent testing.

Description

Technical Field

[0001] This invention relates to the field of sintering furnace monitoring technology, and in particular to a method, system, equipment, and medium for detecting lithium iron phosphate sintering furnaces. Background Technology

[0002] A sintering furnace is a specialized piece of equipment that uses sintering to obtain the desired physical, mechanical properties and microstructure of pressed powder. The sintering temperature plays a crucial role in the sintering process of lithium iron phosphate. In actual production, the temperature control of the sintering furnace must strictly follow the temperature curve of the production process, which means producing according to the "formula". Otherwise, it will lead to the generation of scrap, resulting in great waste and economic losses.

[0003] There are two problems with the existing technology:

[0004] (1) Even when strictly following the temperature curve of the production process, defective products still occurred. The reason for the defective products was that the raw material for the sintering furnace was lithium iron phosphate powder. Different batches of lithium iron phosphate powder came from different sources, and their corresponding processes and raw materials were not completely the same. As a result, the lithium iron phosphate produced according to the original production process curve did not meet the requirements. At the same time, the sintered lithium iron phosphate needed to be packaged into batteries and subjected to coulombic testing to determine whether it was qualified, which wasted a lot of manpower and resources.

[0005] (2) The temperature control of the sintering furnace must strictly follow the temperature curve of the production process; otherwise, it will lead to the generation of scrap, resulting in great waste and economic losses. However, temperature has the characteristics of large inertia, time-varying and serious hysteresis. If the temperature control accuracy is poor, resulting in the sintering temperature being too high or too low, lithium iron phosphate will be scrapped. However, most existing sintering furnaces use PID controllers or fuzzy PID controllers to achieve temperature control. The PID controller uses a very precise thermodynamic model to control the temperature, that is, the accuracy of temperature control depends on the accuracy of the pre-established model. However, this state only exists in an ideal state. Under different conditions such as feeding weight, room temperature, and oxygen content, the pre-designed model cannot accurately judge. Furthermore, in order to improve this shortcoming, existing technologies often use fuzzy PID control, neural network control and other methods to improve the problems of existing technologies. However, multiple experiments have found that simply using data to train the model is prone to overfitting and large deviations. At the same time, the training is random, so the results are not exactly the same each time. Meanwhile, the internal workings of the sintering furnace are quite complex. The sintering process involves parameters such as pressure, oxygen content, and temperature, and each of these parameters is highly correlated. For example, as temperature and pressure rise during sintering, and oxygen content also increases, so does temperature. Furthermore, temperature control suffers from a significant lag; for instance, if the current temperature is 700°C, it cannot be immediately increased to 800°C. These complexities further pose new challenges to the precise control of the algorithm. Summary of the Invention

[0006] In view of this, the purpose of this invention is to provide a method, system, equipment and medium for testing lithium iron phosphate sintering furnaces, which can determine whether the material is qualified from the pressure parameters during the sintering process.

[0007] According to one aspect of the present invention, a method for detecting lithium iron phosphate sintering furnaces is provided, the method comprising:

[0008] Obtain pressure data from historical sintering processes to construct at least one pressure range;

[0009] Collect pressure data at at least one isothermal state during the sintering process;

[0010] By comparing the pressure data with the pressure range, if there is an abnormality, it indicates that there is an abnormality in the sintering material or the sintering process; if there is no abnormality, it indicates that the sintering material is normal.

[0011] In the aforementioned technical solution, different batches of lithium iron phosphate powder originate from different sources, and their corresponding processes and raw materials are not entirely the same, resulting in lithium iron phosphate produced according to the original production process curve failing to meet requirements. Therefore, this case involves monitoring pressure parameters during the sintering process and comparing these parameters with historical sintering pressure data. If the pressure is higher or lower than the specified range, it indicates an anomaly in the batch of material, possibly due to problems with the raw materials or the raw material production process.

[0012] In some embodiments, the at least one pressure range includes a first pressure range, a second pressure range, and a third pressure range;

[0013] The at least one isothermal pressure data includes first isothermal pressure data, second isothermal pressure data, and third isothermal pressure data.

[0014] In the aforementioned technical solution, the temperature curve of the production process includes seven states, in sequence: first heating state, first isothermal state, second heating state, second isothermal state, third heating state, third isothermal state, and cooling state. The pressure values ​​in the first heating state, second heating state, third heating state, and cooling state increase with the temperature, and the pressure value within each state changes over time. Therefore, the pressure values ​​in these four states cannot be accurately determined for monitoring purposes. In view of this, this application selects pressure data from the first isothermal state, second isothermal state, and third isothermal state for monitoring.

[0015] In some embodiments, the points where the isothermal pressure data is taken are:

[0016] The average pressure value during sintering at a constant temperature.

[0017] and / or

[0018] Pressure value at the midpoint of sintering time under isothermal conditions.

[0019] and / or

[0020] The pressure values ​​were randomly obtained over multiple time periods under constant temperature conditions and then averaged.

[0021] In the above technical solution, as mentioned earlier, the temperature control of the sintering furnace must strictly follow the temperature curve of the production process; otherwise, it will lead to the generation of scrap, resulting in significant waste and economic losses. However, temperature has the characteristics of large inertia, time-varying, and severe hysteresis. If the temperature control accuracy is poor, resulting in excessively high or low sintering temperatures, both will lead to the scrapping of lithium iron phosphate. In view of this, in order to ensure that the pressure data in each state can accurately reflect the normal pressure value in that period, the technical solution adopts multiple pressure value selection schemes, which can effectively ensure that the measured data can accurately reflect the normal pressure value in that state.

[0022] In some embodiments, pressure data is compared with a pressure range. If an anomaly is found, it indicates an abnormality in the sintered material or a problem in the sintering process; if no anomaly is found, it indicates that the sintered material is normal. Specifically:

[0023] The first isothermal pressure data is compared with the first pressure range; if it exceeds the range, it is marked as abnormal.

[0024] The second isothermal pressure data is compared with the second pressure range; if it exceeds the range, it is marked as abnormal.

[0025] The third isothermal pressure data is compared with the third pressure range; if it exceeds the range, it is marked as abnormal.

[0026] If the number of abnormalities is ≥2, it indicates that there is an abnormality in the sintering material or an abnormality in the sintering process.

[0027] Otherwise, it indicates that no abnormality has occurred.

[0028] In the above technical solution, this case uses pressure data from three isothermal states—the first, second, and third—for monitoring. Since different batches of lithium iron phosphate powder have different sources, their corresponding processes and raw materials are not entirely the same. Therefore, abnormal situations may occur in different isothermal states, possibly in the first, second, or third isothermal states, or in the first, second, and third isothermal states, or in the first, second, and third isothermal states. The occurrence of an abnormality does not necessarily mean that the lithium iron phosphate produced after sintering from that batch of raw materials is substandard. Therefore, this technical solution further distinguishes between abnormal states. If an abnormality occurs in two states, it indicates a problem with the batch of raw materials. If an abnormality occurs in only one state, further production and packaging into batteries can be carried out for testing to determine if there is an abnormality. This technical solution avoids unnecessary waste.

[0029] In some embodiments, collecting pressure data at at least one isothermal state during sintering, prior to which:

[0030] Establish parameter limit boundaries within the sintering furnace, and set sintering furnace protection strategies based on these parameter limit boundaries;

[0031] Based on the sintering furnace protection strategy, an empirical control rule strategy is set using in-furnace parameters.

[0032] Based on empirical control rules and sintering furnace protection strategies, a fuzzy neural network controller and a PID controller are set up, and a sintering furnace controller is constructed through a Nash equilibrium strategy.

[0033] The temperature parameters of the sintering furnace are adjusted by optimizing the PID parameters of the sintering furnace controller.

[0034] In the aforementioned technical solutions, temperature exhibits significant inertia, time-varying characteristics, and severe hysteresis, leading to poor control accuracy in existing methods. Poor temperature control accuracy can result in excessively high or low sintering temperatures during temperature adjustment, causing lithium iron phosphate to fail. To ensure the accuracy and stability of pressure data detection and to correctly identify abnormal situations, this invention proposes a new applicable method for temperature control during sintering. During sintering, changes in various parameters cause global changes, which can be understood as each parameter being a player in a game of temperature control. Therefore, this invention introduces the concept of game theory, which studies how multiple decision-making agents interact to choose the appropriate decision to maximize overall benefit, i.e., reaching an equilibrium state called Nash equilibrium. In this invention, this corresponds to how to set each parameter to achieve the best overall temperature control effect during the temperature control process. Furthermore, the solution value of the Nash equilibrium is input into a fuzzy neural network controller to further correct the parameter values, thereby ensuring the accuracy of data control. This invention performs a first round of data screening based on Nash equilibrium, then uses a fuzzy neural network to further refine and screen the screened values, and finally inputs them into a PID controller for further control, thereby achieving the best overall temperature regulation effect.

[0035] In some embodiments, establishing parameter limit boundaries within the sintering furnace and setting a sintering furnace protection strategy based on these parameter limit boundaries specifically includes the following steps:

[0036] Obtain the boundaries of oxygen content, pressure, temperature, and sintering time in the furnace; determine the protection strategy for the sintering furnace based on these four boundaries.

[0037] In the above technical solution, four process parameters that have the greatest impact on the sintering process were selected to construct the safety boundary. The purpose is to ensure that the scrap rate of sintering can be kept within a certain range under the condition of model deviation, so as not to lead to the scrapping of the entire furnace.

[0038] In some embodiments, the empirical control rule strategy based on sintering furnace protection strategy and using furnace parameters to set the strategy specifically includes the following steps:

[0039] When the temperature is below the first temperature threshold, the furnace is in a heating state, and the furnace temperature is controlled.

[0040] When the temperature equals the first temperature threshold, the furnace is in a first isothermal state, and the furnace temperature is controlled to stop rising.

[0041] If the time is greater than the first time threshold but less than the second time threshold, the furnace transitions from the first isothermal state to the second isothermal state, thus controlling the temperature rise inside the furnace.

[0042] When the temperature equals the second temperature threshold, the furnace is in a second isothermal state, and the furnace temperature is controlled to stop rising.

[0043] When the time is greater than the second time threshold and less than the third time threshold, the furnace transitions from the second isothermal state to the third isothermal state, thus controlling the temperature rise inside the furnace.

[0044] When the temperature equals the third temperature threshold, the furnace is in the third isothermal state, and the furnace temperature is controlled to stop rising.

[0045] When the temperature exceeds the third threshold, the furnace transitions from the third isothermal state to a cooling state.

[0046] In the aforementioned technical solutions, traditional sintering furnaces often use the temperature curve of the production process for sintering. Therefore, this case needs to consider the importance of historical data. Thus, the empirical control rule strategy is formulated based on the temperature curve of the production process. Its purpose is to ensure that, even with model deviations, the scrap rate of sintering can be kept within a certain range, preventing the entire furnace from being scrapped. Simultaneously, utilizing past and empirical data helps provide valuable reference during the Nash equilibrium solution process and can improve the accuracy of the model in subsequent neural network model training.

[0047] In some embodiments, a fuzzy neural network controller and a PID controller are set based on an empirical control rule strategy and a sintering furnace protection strategy, and a sintering furnace controller is constructed using a Nash equilibrium strategy:

[0048] Establish a strategy set and a benefit set based on empirical data of temperature under empirical control rules and sintering furnace protection strategies.

[0049] The Nash equilibrium solution is obtained by solving the payoff set, and the Nash equilibrium set is constructed based on the solution value.

[0050] The optimal set is obtained by optimizing the fuzzy neural network controller based on the input of the Nash equilibrium set;

[0051] The optimal set is input into the PID controller to construct the sintering furnace controller.

[0052] In the above technical solution, the present invention performs the first round of data screening from Nash equilibrium, and then uses a fuzzy neural network to further refine and screen the screened values, and then inputs them into a PID controller for further control, so as to achieve the best overall temperature control effect.

[0053] In some embodiments, the optimal Nash equilibrium set is obtained by optimizing the fuzzy neural network controller based on the input of the Nash equilibrium set. Specifically:

[0054] When the Nash equilibrium set has a unique solution, the unique solution is optimized using a fuzzy neural network.

[0055] When there are infinitely many solutions to the Nash equilibrium set, the combination with the highest return in the Nash equilibrium set is selected and input into the fuzzy neural network for optimization and filtering.

[0056] When the Nash equilibrium set is empty, the combination with the highest return in the payoff set is selected and input into the fuzzy neural network for optimization.

[0057] In the above technical solution, there are three cases of Nash equilibrium solution: (1) unique solution (2) infinitely many solutions (3) no solution. Therefore, in order to avoid invalid data screening that prevents the PID controller from receiving control commands, the present invention further sets the above three states, which can effectively ensure that the PID controller can receive data regulation and control the sintering furnace.

[0058] In some embodiments, the strategy set is a set of combinations of parameter values ​​under temperature changes under the sintering furnace protection strategy;

[0059] The payout set is constructed by using different payout values ​​to determine the qualification criteria of powder obtained from various combinations of values ​​in the strategy set, and then constructing the payout set based on these payout values.

[0060] In the above technical solution, the objective is precise temperature control. Therefore, temperature can be understood as the decision-making agent in the game, and the decision made by this agent has a significant impact on the control of the furnace temperature. It is important to note that the temperature here refers to the set temperature parameter, not the real-time furnace temperature. Furthermore, the benefit can be determined based on the medium particle size obtained from sintering using historical data. The medium particle size of typical sintered lithium iron phosphate is generally between 3-8 μm (D50). Using 5 μm as a standard, the benefit score for different combinations of medium particle sizes can be defined. It should be understood that, as described in the background, the sintering furnace is complex. If a thermodynamic model of the sintering furnace is used to establish the benefit function and build the benefit set, numerous complex calculations are required. Therefore, this solution uses the qualified standard for particle size in lithium iron phosphate to build the benefit set. It should also be clear that other methods such as tap density, specific surface area, and electrical conductivity can also be used for judgment. The reason why particle size is used as the evaluation standard in this case is that during the sintering process, temperature and sintering time have the greatest impact on the particle size of the powder. In other words, it can be understood that particle size and temperature are highly correlated.

[0061] In some embodiments, as an alternative implementation, the benefit set is constructed by taking the heat loss ratio under each combination of values ​​in the strategy set as the standard, and the benefit set is constructed by the benefit values.

[0062] In the above technical solution, compared to the previous implementation scheme, the heat loss ratio is used as the criterion for determining the benefit set. This is because traditional sintering furnace temperature control models are mostly based on mathematical models established according to the sintering furnace temperature. The three important parameters in these models are the system output temperature, the system input temperature, and the system loss temperature. Ideally, the furnace temperature during sintering should strictly follow the temperature curve. However, if the system loss increases, the temperature deviation inside the furnace will also increase. Therefore, this proposal further suggests an alternative benefit set establishment scheme based on the heat loss ratio. A higher heat loss indicates instability during sintering and a greater probability of producing defective products.

[0063] According to another aspect of the present invention, a detection system for a lithium iron phosphate sintering furnace is provided, the system comprising:

[0064] The acquisition module, data collection module, and judgment module are electrically connected in sequence.

[0065] Acquisition module: used to acquire pressure data from historical sintering processes to construct at least one pressure range;

[0066] Acquisition module: Used to acquire pressure data at at least one isothermal state during the sintering process;

[0067] Judgment module: The pressure data is compared with the pressure range. If the pressure exceeds the range, it indicates that there is an abnormality in the sintering material or the sintering process. If the pressure does not exceed the range, it indicates that the sintering material is normal.

[0068] In the aforementioned technical solution, different batches of lithium iron phosphate powder originate from different sources, and their corresponding processes and raw materials are not entirely the same, resulting in lithium iron phosphate produced according to the original production process curve failing to meet requirements. Therefore, this system utilizes various modules to monitor pressure parameters during the sintering process and compares these pressure parameters with historical sintering pressure data. If the pressure is higher or lower than the specified range, it indicates an anomaly in the batch of material, possibly due to problems with the raw materials or the raw material production process.

[0069] In some embodiments, the system further includes a control module electrically connected to the acquisition module, used to establish parameter limit boundaries within the sintering furnace, set a sintering furnace protection strategy based on the parameter limit boundaries; set an empirical control rule strategy using the furnace parameters based on the sintering furnace protection strategy; set a fuzzy neural network controller and a PID controller based on the empirical control rule strategy and the sintering furnace protection strategy, and construct a sintering furnace controller through a Nash equilibrium strategy; and optimize the PID parameters to adjust the sintering furnace temperature parameters based on the sintering furnace controller.

[0070] The above technical solution addresses the issue that changes in any parameter will cause global changes, which can be understood as each parameter being a player in a game of temperature control. Therefore, this invention introduces game theory, which studies how multiple decision-making agents interact to choose the appropriate decision to maximize overall benefit, i.e., to reach an equilibrium state called Nash equilibrium. In this invention, this corresponds to how to set each parameter to achieve the best overall temperature control effect during the temperature control process. The solution value of Nash equilibrium is further used as input to a fuzzy neural network controller to further refine the parameter values, thereby ensuring the accuracy of data control. This invention performs a first round of data filtering based on Nash equilibrium, then uses a fuzzy neural network to further refine and filter the filtered values ​​before inputting them into a PID controller for further control, thus achieving the best overall temperature control effect.

[0071] According to another aspect of the present invention, a testing device for a lithium iron phosphate sintering furnace is provided, comprising:

[0072] At least one processor, and a memory communicatively connected to the processor, wherein,

[0073] The memory stores instructions that can be executed by at least one processor, and when the instructions are executed by at least one of the processors, the above-described method for detecting a lithium iron phosphate sintering furnace is performed.

[0074] According to another aspect of the present invention, a computer-readable storage medium is provided, storing a computer program that, when executed by a processor, implements the above-described method for detecting a lithium iron phosphate sintering furnace. Attached Figure Description

[0075] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0076] Figure 1 This is a schematic flowchart of an embodiment of a lithium iron phosphate sintering furnace detection method according to the present invention;

[0077] Figure 2 It is a sintering process curve commonly used in existing technologies;

[0078] Figure 3 This is a schematic diagram of the S21-S24 process of an embodiment of the lithium iron phosphate sintering furnace detection method of the present invention;

[0079] Figure 4This is a schematic diagram of steps S21-S24 of an embodiment of the lithium iron phosphate sintering furnace detection method of the present invention.

[0080] Figure 5 This is a schematic diagram of the system framework of an embodiment of the lithium iron phosphate sintering furnace detection system of the present invention;

[0081] Figure 6 This is a schematic diagram of the system framework of another embodiment of the lithium iron phosphate sintering furnace detection system of the present invention. Detailed Implementation

[0082] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be particularly noted that the following embodiments are for illustrative purposes only and do not limit the scope of the invention. Similarly, the following embodiments are only some, not all, embodiments of the present invention, and all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0083] This invention proposes a testing method for lithium iron phosphate sintering furnaces, which can determine whether the material is qualified based on the pressure parameters during the sintering process.

[0084] Please see Figure 1 , Figure 1 This is a schematic flowchart of an embodiment of a lithium iron phosphate sintering furnace testing method according to the present invention; it should be noted that if substantially the same results are obtained, the method of the present invention is not necessarily identical. Figure 1 The sequence of processes shown is limited.

[0085] S1 acquires pressure data from historical sintering processes to construct at least one pressure range;

[0086] It is important to note that the historical pressure data from the sintering process is quite extensive. When establishing the pressure range, the pressure data should first be sorted from smallest to largest, and extreme values ​​at the beginning and end should be filtered out. This means removing several maximum and minimum values, and finally establishing the pressure range based on the minimum and maximum values ​​from the filtered set. This approach aims to avoid the influence of extreme values ​​on the overall assessment. While some materials may sinter into acceptable lithium iron phosphate materials under extreme conditions, this also means that the raw material is fluctuating between acceptable and unacceptable levels. To be on the safe side, further filtering of these marginal values ​​is necessary when selecting appropriate values.

[0087] S2 collects pressure data at at least one isothermal state during the sintering process;

[0088] The at least one pressure range includes a first pressure range, a second pressure range, and a third pressure range; the at least one isothermal pressure data includes first isothermal pressure data, second isothermal pressure data, and third isothermal pressure data.

[0089] In the above technical solutions, please refer to Figure 2 The temperature curve of the production process includes seven states, in sequence: first heating state, first isothermal state, second heating state, second isothermal state, third heating state, third isothermal state, and cooling state. The pressure values ​​in the first heating state, second heating state, third heating state, and cooling state increase with the temperature, and the pressure value within each state changes over time. Therefore, the pressure values ​​in these four states cannot be accurately determined for monitoring purposes. In view of this, this case uses pressure data from the first isothermal state, second isothermal state, and third isothermal state for monitoring. It should be noted that different production process curves may have fewer than three isothermal states or more than three isothermal states; this embodiment will not elaborate further. The key is to ensure that the pressure values ​​under the isothermal state remain stable.

[0090] Furthermore, the isothermal pressure data is collected from the following points: the average pressure value during the sintering time under isothermal conditions, and / or the pressure value at the midpoint of the sintering time under isothermal conditions, and / or the average of randomly acquired pressure values ​​over multiple time periods under isothermal conditions. As mentioned earlier, the temperature control of the sintering furnace must strictly follow the temperature curve of the production process; otherwise, it will lead to the generation of scrap, resulting in significant waste and economic losses. However, temperature has the characteristics of large inertia, time-varying, and severe hysteresis. If the temperature control accuracy is poor, resulting in excessively high or low sintering temperatures, lithium iron phosphate will be scrapped. In view of this, in order to ensure that the pressure data in each state can correctly reflect the normal pressure value in that period, the technical solution adopts multiple pressure value collection schemes, which can effectively ensure that the measured data can correctly reflect the normal pressure value in that state.

[0091] Furthermore, S2 collects pressure data at at least one isothermal state during the sintering process, and previously includes: S21, establishing parameter limit boundaries within the sintering furnace, and setting a sintering furnace protection strategy based on the parameter limit boundaries;

[0092] S22 is based on the sintering furnace protection strategy and uses the furnace parameters to set empirical control rules.

[0093] S23 is based on empirical control rules and sintering furnace protection strategies. It sets up a fuzzy neural network controller, a PID controller, and constructs a sintering furnace controller through a Nash equilibrium strategy.

[0094] S24 optimizes PID parameters based on the sintering furnace controller to adjust the sintering furnace temperature parameters.

[0095] Since changes in each parameter cause global changes, each parameter can be understood as a player in a game of temperature control. Therefore, this invention introduces the concept of game theory, which studies how multiple decision-making agents interact to choose the appropriate decision to maximize overall benefit, i.e., to reach an equilibrium state called Nash equilibrium. In this invention, this corresponds to how to set each parameter to achieve the best overall temperature control effect during the temperature control process. The solution value of the Nash equilibrium is further used as input to a fuzzy neural network controller to further refine the parameter values, thereby ensuring the accuracy of data control. This invention performs a first round of data filtering based on the Nash equilibrium, then uses a fuzzy neural network to further refine and filter the filtered values ​​before inputting them into a PID controller for further control, thus achieving the best overall temperature control effect.

[0096] Please see Figure 3 , Figure 4 , Figure 3 This is a schematic diagram of the S21-S24 process of an embodiment of the lithium iron phosphate sintering furnace detection method of the present invention; Figure 4 This is a schematic diagram of steps S21-S24 of an embodiment of the lithium iron phosphate sintering furnace detection method of the present invention. It should be noted that if substantially the same result is obtained, the method of the present invention does not necessarily require further clarification. Figure 3 The sequence of processes shown is limited.

[0097] like Figure 3 As shown, the method includes the following steps:

[0098] S21 establishes parameter limit boundaries within the sintering furnace and sets sintering furnace protection strategies based on these parameter limit boundaries;

[0099] In this embodiment, this step specifically requires obtaining the boundaries of oxygen content, pressure, temperature, and sintering time within the furnace; based on these four boundaries, a protection strategy for the sintering furnace is determined. Four process parameters with the greatest impact on the sintering process are selected to construct the safety boundaries. The purpose is to ensure that, given model deviations, the scrap rate of sintering remains within a certain range, preventing the entire furnace from being scrapped. It should be understood that the parameters selected in this embodiment are generally applicable to various sintering furnaces. If individual sintering furnaces have additional important control parameters, these should also be included as limiting boundaries during the modeling process, such as the weight of powder and pressed blanks, the appearance of pressed blanks, and the geometric dimensions and deviations of pressed blanks.

[0100] S22 is based on the sintering furnace protection strategy and uses the furnace parameters to set empirical control rules.

[0101] In this embodiment, the following strategies are specifically included:

[0102] When the temperature is below the first temperature threshold, the furnace is in a heating state, and the furnace temperature is controlled.

[0103] When the temperature equals the first temperature threshold, the furnace is in a first isothermal state, and the furnace temperature is controlled to stop rising.

[0104] If the time is greater than the first time threshold but less than the second time threshold, the furnace transitions from the first isothermal state to the second isothermal state, thus controlling the temperature rise inside the furnace.

[0105] When the temperature equals the second temperature threshold, the furnace is in a second isothermal state, and the furnace temperature is controlled to stop rising.

[0106] When the time is greater than the second time threshold and less than the third time threshold, the furnace transitions from the second isothermal state to the third isothermal state, thus controlling the temperature rise inside the furnace.

[0107] When the temperature equals the third temperature threshold, the furnace is in the third isothermal state, and the furnace temperature is controlled to stop rising.

[0108] When the temperature exceeds the third threshold, the furnace transitions from the third isothermal state to a cooling state.

[0109] In this embodiment, traditional sintering furnaces typically use the temperature profile of the production process for sintering. Please refer to [link to relevant documentation]. Figure 3 Existing sintering curves mostly use graphical curves, sequentially including: first heating state → first isothermal state → second heating state → second isothermal state → third heating state → third isothermal state → cooling state. The difference between different process curves lies in the heating rate, heating time, isothermal temperature, isothermal time, and cooling time in different states. These parameters are process parameters with high yield rates summarized from past experience; therefore, this case needs to consider the importance of historical data. In this embodiment, the empirical control rule strategy is formulated based on the temperature curve of the production process. Its purpose is to ensure that, even with model deviations, the scrap rate of sintering can be kept within a certain range, without leading to the scrapping of the entire furnace. At the same time, using past data and empirical data helps to provide certain reference value in the Nash equilibrium solution process and can also improve the accuracy of the model in the subsequent training of the neural network model. It should also be noted that the above thresholds correspond to the inflection points before each state in the figure, which will not be explained in detail here.

[0110] S23 is based on empirical control rules and sintering furnace protection strategies. It sets up a fuzzy neural network controller, a PID controller, and constructs a sintering furnace controller through a Nash equilibrium strategy.

[0111] In this embodiment, the specific steps include the following:

[0112] S231 establishes a strategy set and a benefit set based on empirical control rules and strategies and empirical temperature data under sintering furnace protection strategies.

[0113] The strategy set is a collection of combinations of parameter values ​​under temperature changes under the sintering furnace protection strategy; a detailed description is as follows:

[0114] In this embodiment, four players participate in the game: the oxygen content in the furnace, the pressure in the furnace, the temperature in the furnace, and the sintering time, denoted as q1, q2, q3, and q4, respectively. Each player has a set of strategies to choose from, namely D1, D2, D3, and D4, and the number of selectable strategies in each strategy set is n1...n. n The strategies adopted by the four players in this game are d1∈D1, d2∈D2, d3∈D3, and d4∈D4.

[0115] The payoff set is constructed by setting different payoff values ​​for the qualified standards of powder obtained from each combination of values ​​in the strategy set, and the payoff set is constructed by these payoff values. The expression for a set of payoffs corresponding to a set of strategies adopted by the players is as follows:

[0116] f(d1,d2,d3,d4)=(b1,b2,b3,b4)

[0117] All the combinations of returns form return set B.

[0118] In this embodiment, a medium particle size (D50) of 3-8 μm is used as the standard. The particle size benefit function is as follows:

[0119]

[0120] It should be noted that the above profit function b i This is a simplified formula used in this case for ease of explanation and understanding, and is not intended as the only valid definition. The purpose of this case is precise temperature control; therefore, it can be understood that temperature is a decision-making agent in the game process, and the agent's chosen decision has a significant impact on the control of the furnace temperature. It is important to note that the temperature here refers to the set temperature parameter, not the real-time furnace temperature. Furthermore, the payoff can be determined based on the medium particle size obtained from sintering using historical data. The typical medium particle size (D50) of sintered lithium iron phosphate is 3-8 μm. Using 5 μm as a standard, the payoff score for different combinations of medium particle sizes can be defined. It is important to understand that, as described in the background technology of this case, the sintering furnace is quite complex. If a thermodynamic model of the sintering furnace is used to establish the payoff function to build the payoff set, many complex calculations would be required. Therefore, this case uses the qualified standard for particle size in lithium iron phosphate to establish the payoff set. It should also be clear that other methods such as tap density, specific surface area, and electrical conductivity can also be used for judgment. The reason why particle size is used as the evaluation standard in this case is that during the sintering process, temperature and sintering time have the greatest impact on the particle size of the powder. In other words, it can be understood that particle size and temperature are highly correlated.

[0121] Furthermore, as an alternative implementation, the revenue set is constructed by using the heat loss ratio as a standard for each combination of values ​​in the strategy set, and then constructing the revenue set based on these revenue values. The formula is as follows:

[0122]

[0123] In the formula, q is the proportion of heat loss, Q 损 Q represents the amount of heat loss from the system output. 总 This refers to the heat output by the system.

[0124]

[0125] It should be noted that the above profit function b i This is a simplified formula used in this case for ease of explanation and understanding, and is not intended as the only limitation. In the above technical solution, compared to the previous implementation scheme, the heat loss ratio is used here to determine the benefit set. This is because traditional sintering furnace temperature control models are mostly based on mathematical models established according to the sintering furnace temperature, and the three important parameters in this mathematical model are the system output temperature, the system input temperature, and the system loss temperature. Ideally, the furnace temperature changes strictly according to the temperature curve during the sintering process. If the system loss increases, the temperature deviation in the furnace will increase. In view of this, this case further proposes an alternative benefit set establishment scheme, which is the heat loss ratio. The greater the heat loss, the more unstable the sintering process is, and the greater the probability of producing defective products.

[0126] S232 solves for the Nash equilibrium solution by using the payoff set and constructs the Nash equilibrium set based on the solution value;

[0127] A Nash equilibrium is a set of strategies such that when a player adopts this set of strategies, changing their own strategy will not increase their payoff. Therefore, for any d... 1i ∈D1、d 2i ∈D2、d 3i ∈D3、d 4i ∈D4, i∈1……n. Nash equilibrium set The following conditions must be met:

[0128]

[0129]

[0130] Where, d 1i …d 2i …d 3i …d 4iThis represents a player choosing the i-th strategy from the corresponding strategy set example. From the above formula, we can see that Nash equilibrium solutions exist in three states: unique solution, infinitely many solutions, and no solution.

[0131] S233 optimizes the fuzzy neural network controller based on the Nash equilibrium set input to obtain the optimal set; specifically, it consists of the following three steps:

[0132] When the Nash equilibrium set has a unique solution, the unique solution is optimized using a fuzzy neural network.

[0133] When there are infinitely many solutions to the Nash equilibrium set, the combination with the highest return in the Nash equilibrium set is selected and input into the fuzzy neural network for optimization and filtering.

[0134] When the Nash equilibrium set is empty, the combination with the highest return in the payoff set is selected and input into the fuzzy neural network for optimization.

[0135] In the above technical solution, as described in S32, there are three cases of Nash equilibrium solution: (1) unique solution (2) infinitely many solutions (3) no solution. Therefore, in order to avoid invalid data screening that prevents the PID controller from receiving control commands, the present invention further sets the above three states, which can effectively ensure that the PID controller can receive data regulation and control the sintering furnace.

[0136] Furthermore, fuzzy neural network controllers and PID controllers are both existing technologies and will not be elaborated upon here. It is understandable that as long as a structure can be constructed... Figure 2 The architecture shown is sufficient. Please refer to [link / reference]. Figure 2 Because the operating mechanism of the sintering furnace is complex and the system has multivariable and nonlinear characteristics, after the Nash equilibrium ensemble output, a PID controller is combined with a fuzzy neural network. The fuzzy neural network combines the advantages of high accuracy in fuzzy control and strong self-learning ability in neural networks. A control scheme is designed with the sintering furnace as the controlled object. The designed Nash equilibrium strategy and the fuzzy neural network PID controller structure are as follows: Figure 2 As shown. In Figure 2 In the middle, A i P i T i S i The corresponding oxygen content, pressure, temperature, and sintering time in the furnace represent the desired output of the sintering furnace; A j P j T j S jThis corresponds to the actual output of the sintering furnace. The input to the PID controller is the deviation between the expected output and the actual output after solving the Nash equilibrium set. The input to the fuzzy neural network is the deviation between the expected output and the actual output and the rate of change of the deviation after solving the Nash equilibrium set; the output of the fuzzy neural network is the four optimal control parameters ΔA of the sintering furnace PID controller after online learning and optimization. i ΔP i ΔT i ΔS i The above technical solution is an existing technical architecture, and will not be further described in this embodiment.

[0137] S234 inputs the optimal set into the PID controller to construct the sintering furnace controller.

[0138] In this embodiment, the idea of ​​step S3 is to use Nash equilibrium to perform the first round of data screening, and then use fuzzy neural network to further refine and screen the screened values, and then input the data into the PID controller for further control, so as to achieve the best overall temperature control effect.

[0139] S24 optimizes PID parameters based on the sintering furnace controller to adjust the sintering furnace temperature parameters.

[0140] In the above embodiments, since changes in each parameter cause global changes, each parameter can be understood as a player in a game of temperature control. Therefore, this invention introduces the concept of game theory, which studies how multiple decision-making agents interact to choose appropriate decisions to maximize overall gains, i.e., to reach an equilibrium state called Nash equilibrium. In this invention, this corresponds to how to set each parameter to achieve the best overall temperature control effect during the temperature control process. The solution value of the Nash equilibrium is further used as input to the fuzzy neural network controller to further refine the parameter values, thereby ensuring the accuracy of data control. This invention performs a first round of data filtering based on the Nash equilibrium, then uses the fuzzy neural network to further refine and filter the filtered values ​​before inputting them into the PID controller for further control, thus achieving the best overall temperature control effect.

[0141] S3 compares the pressure data with the pressure range. If there is an abnormality, it indicates that there is an abnormality in the sintering material or the sintering process; if there is no abnormality, it indicates that the sintering material is normal.

[0142] Specifically, the first isothermal pressure data is compared with the first pressure range, and if it exceeds the range, it is marked as abnormal;

[0143] The second isothermal pressure data is compared with the second pressure range; if it exceeds the range, it is marked as abnormal.

[0144] The third isothermal pressure data is compared with the third pressure range; if it exceeds the range, it is marked as abnormal.

[0145] If the number of abnormalities is ≥2, it indicates that there is an abnormality in the sintering material or an abnormality in the sintering process.

[0146] Otherwise, it indicates that no abnormality has occurred.

[0147] In the above technical solution, this case uses pressure data from three isothermal states—the first, second, and third—for monitoring. Since different batches of lithium iron phosphate powder have different sources, their corresponding processes and raw materials are not entirely the same. Therefore, abnormal situations may occur in different isothermal states, possibly in the first, second, or third isothermal states, or in the first, second, and third isothermal states, or in the first, second, and third isothermal states. The occurrence of an abnormality does not necessarily mean that the lithium iron phosphate produced after sintering from that batch of raw materials is substandard. Therefore, this technical solution further distinguishes between abnormal states. If an abnormality occurs in two states, it indicates a problem with the batch of raw materials. If an abnormality occurs in only one state, further production and packaging into batteries can be carried out for testing to determine if there is an abnormality. This technical solution avoids unnecessary waste.

[0148] According to another aspect of the present invention, an embodiment of a lithium iron phosphate sintering furnace detection system is provided; please refer to [link to relevant documentation]. Figure 5 The system includes:

[0149] The acquisition module, data collection module, and judgment module are electrically connected in sequence.

[0150] Acquisition module: used to acquire pressure data from historical sintering processes to construct at least one pressure range;

[0151] Acquisition module: Used to acquire pressure data at at least one isothermal state during the sintering process;

[0152] Judgment module: The pressure data is compared with the pressure range. If the pressure exceeds the range, it indicates that there is an abnormality in the sintering material or the sintering process. If the pressure does not exceed the range, it indicates that the sintering material is normal.

[0153] In the aforementioned technical solution, different batches of lithium iron phosphate powder originate from different sources, and their corresponding processes and raw materials are not entirely the same, resulting in lithium iron phosphate produced according to the original production process curve failing to meet requirements. Therefore, this system utilizes various modules to monitor pressure parameters during the sintering process and compares these pressure parameters with historical sintering pressure data. If the pressure is higher or lower than the specified range, it indicates an anomaly in the batch of material, possibly due to problems with the raw materials or the raw material production process.

[0154] Further, please refer to Figure 6 The system also includes a control module electrically connected to the acquisition module, used to establish parameter limit boundaries within the sintering furnace, set sintering furnace protection strategies based on these parameter limit boundaries, set empirical control rule strategies based on the furnace parameters based on the sintering furnace protection strategies, set fuzzy neural network controllers and PID controllers based on the empirical control rule strategies and the sintering furnace protection strategies, and construct a sintering furnace controller through a Nash equilibrium strategy, and optimize the PID parameters to adjust the sintering furnace temperature parameters based on the sintering furnace controller.

[0155] The above technical solution addresses the issue that changes in any parameter will cause global changes, which can be understood as each parameter being a player in a game of temperature control. Therefore, this invention introduces game theory, which studies how multiple decision-making agents interact to choose the appropriate decision to maximize overall benefit, i.e., to reach an equilibrium state called Nash equilibrium. In this invention, this corresponds to how to set each parameter to achieve the best overall temperature control effect during the temperature control process. The solution value of Nash equilibrium is further used as input to a fuzzy neural network controller to further refine the parameter values, thereby ensuring the accuracy of data control. This invention performs a first round of data filtering based on Nash equilibrium, then uses a fuzzy neural network to further refine and filter the filtered values ​​before inputting them into a PID controller for further control, thus achieving the best overall temperature control effect.

[0156] Each module and method step in the above system corresponds to one another, and will not be repeated here.

[0157] According to another aspect of the present invention, an embodiment of a lithium iron phosphate sintering furnace testing device is provided, comprising:

[0158] At least one processor, and a memory communicatively connected to the processor, wherein,

[0159] The memory stores instructions that can be executed by at least one processor, and when the instructions are executed by at least one of the processors, the above-described method for detecting a lithium iron phosphate sintering furnace is performed.

[0160] According to another aspect of the present invention, an embodiment of a computer-readable storage medium is provided, which stores a computer program, characterized in that the computer program, when executed by a processor, implements the above-described method for detecting a lithium iron phosphate sintering furnace.

[0161] In the above embodiments, the present invention performs a first round of data screening based on Nash equilibrium, and then uses a fuzzy neural network to further refine and screen the screened values ​​before inputting them into a PID controller for further control, thereby achieving the best overall temperature control effect.

[0162] The above description is only a part of the embodiments of the present invention and does not limit the scope of protection of the present invention. Any equivalent device or equivalent process transformation made based on the content of the present invention specification and drawings, or direct or indirect application in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A lithium iron phosphate sintering furnace detection method, characterized in that, The method comprises the following steps: acquiring pressure data in a historical sintering process to construct at least one pressure range; acquiring pressure data of at least one constant temperature state in the sintering process; comparing the pressure data with the pressure range, and if abnormal, indicating that the sintering material is abnormal or the sintering process is abnormal; if not abnormal, indicating that the sintering material is normal; acquiring pressure data of at least one constant temperature state in the sintering process, and the method further comprises the following steps: establishing a parameter limit boundary in the sintering furnace, and setting a sintering furnace protection strategy based on the parameter limit boundary, which comprises the following steps: acquiring an oxygen content boundary in the furnace, a pressure boundary in the furnace, a temperature boundary in the furnace, and a sintering time boundary; and determining a sintering furnace protection strategy based on the four boundaries; based on the sintering furnace protection strategy, using an empirical control rule strategy to set the furnace parameters, which comprises the following steps: when the temperature is lower than a first temperature threshold value, the furnace is in a heating state, and the furnace is controlled to heat; when the temperature is equal to the first temperature threshold value, the furnace is in a first constant temperature state, and the furnace is controlled to stop heating; when the time is greater than a first time threshold value and less than a second time threshold value, the furnace is in a first constant temperature state to a second constant temperature state, and the furnace is controlled to heat; when the temperature is equal to a second temperature threshold value, the furnace is in a second constant temperature state, and the furnace is controlled to stop heating; when the time is greater than the second time threshold value and less than a third time threshold value, the furnace is in a second constant temperature state to a third constant temperature state, and the furnace is controlled to heat; when the temperature is equal to a third temperature threshold value, the furnace is in a third constant temperature state, and the furnace is controlled to stop heating; and when the temperature is greater than the third threshold value, the furnace is in a third constant temperature state to a cooling state; based on the empirical control rule strategy and the sintering furnace protection strategy, setting a fuzzy neural network controller and a PID controller and constructing a sintering furnace controller through a Nash equilibrium strategy; based on the sintering furnace controller, optimizing the PID parameter to adjust the temperature parameter of the sintering furnace; based on the empirical control rule strategy and the sintering furnace protection strategy, setting a fuzzy neural network controller and a PID controller and constructing a sintering furnace controller: establishing a strategy set and a benefit set according to the empirical data of the temperature under the empirical control rule strategy and the sintering furnace protection strategy; solving a Nash equilibrium solution through the benefit set, and constructing a Nash equilibrium set according to the solution value; based on the Nash equilibrium set, inputting the fuzzy neural network controller for optimization to obtain an optimal set; inputting the optimal set into the PID controller to construct the sintering furnace controller.

2. The lithium iron phosphate sintering furnace detection method of claim 1, wherein the at least one pressure range comprises a first pressure range, a second pressure range, and a third pressure range; and the at least one constant temperature state pressure data comprises first constant temperature state pressure data, second constant temperature state pressure data, and third constant temperature state pressure data.

3. The lithium iron phosphate sintering furnace detection method of claim 2, wherein the constant temperature state pressure data is obtained at a value point, which is an average pressure value in a sintering time at a constant temperature state, a pressure value at a midpoint of the sintering time at the constant temperature state, or a pressure value randomly obtained in a plurality of time periods at the constant temperature state and then averaged.

4. The lithium iron phosphate sintering furnace detection method of claim 2, wherein ​ ​ ​ ​ If the pressure data is compared with the pressure range and is abnormal, it indicates that the sintered material is abnormal or the sintering process is abnormal; if it is not abnormal, it indicates that the sintered material is normal, and the specific steps are as follows: The first constant temperature pressure data is compared with the first pressure range, and if it exceeds the range, it is marked as abnormal; The second constant temperature pressure data is compared with the second pressure range, and if it exceeds the range, it is marked as abnormal; The third constant temperature pressure data is compared with the third pressure range, and if it exceeds the range, it is marked as abnormal; If the number of abnormalities is greater than or equal to 2, it indicates that the sintered material is abnormal or the sintering process is abnormal; Otherwise, it indicates that no abnormality has occurred.

5. A lithium iron phosphate sintering furnace detection system, characterized in that, The application is applied to the lithium iron phosphate sintering furnace detection method in any one of claims 1-4; the system comprises: sequentially connected acquisition module, collection module and determination module; The acquisition module is used to acquire pressure data in the historical sintering process to construct at least one pressure range; The collection module is used to collect at least one constant temperature pressure data in the sintering process; The determination module compares the pressure data with the pressure range, and if it exceeds the range, it indicates that the sintered material is abnormal or the sintering process is abnormal; if it does not exceed the range, it indicates that the sintered material is normal.

6. The lithium iron phosphate sintering furnace detection system of claim 5, wherein, The system further comprises a control module electrically connected to the collection module, which is used to establish a parameter limit boundary in the sintering furnace, and set a sintering furnace protection strategy based on the parameter limit boundary; Based on the sintering furnace protection strategy, an experience control rule strategy is used; based on the experience control rule strategy and the sintering furnace protection strategy, a fuzzy neural network controller and a PID controller are set and a sintering furnace controller is constructed through a Nash equilibrium strategy; based on the sintering furnace controller, the PID parameter is adjusted to optimize the sintering furnace temperature parameter.

7. A lithium iron phosphate sintering furnace detection device, characterized by, It comprises: at least one processor and a memory connected to the processor in communication, wherein The memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the method for detecting the lithium iron phosphate sintering furnace in any one of claims 1 to 4 is performed.

8. A computer readable storage medium storing a computer program, characterized in that, The computer program is executed by the processor to realize the lithium iron phosphate sintering furnace detection method in any one of claims 1 to 4.

Images