A method and apparatus for predicting atomization in a reduction furnace

Abstract

The application provides a prediction method and device for atomization of a reduction furnace, and relates to the technical field of polysilicon production. The method comprises the following steps: determining a theoretical exhaust temperature of reduction tail gas at a first time according to a prediction model; determining a prediction temperature at a second time according to the actual exhaust temperature of the reduction tail gas at the first time and the theoretical exhaust temperature of the reduction tail gas; the second time is a time after the first time; and predicting the atomization reaction of the reduction furnace at the second time according to the prediction temperature. According to the scheme of the application, the atomization degree of the reduction furnace can be accurately judged according to the prediction temperature, so that the operator can timely master the atomization degree in the reduction furnace and make adjustment, the reduction furnace can be operated at the boundary of a certain degree of atomization, and the purpose of reducing power consumption and ensuring the product quality of the reduction furnace can be achieved.

Description

Technical Field

[0001] This invention relates to the field of polysilicon production technology, and in particular to a method and apparatus for predicting atomization in a reduction furnace. Background Technology

[0002] During the 110-hour polysilicon production process, how to promptly sense the furnace status and make adjustments in advance to prevent abnormal phenomena such as atomization is currently the most pressing issue in this field. However, as a gray box system, the surface temperature of the silicon rod is difficult to obtain directly through sensors. On-site personnel often use some indirect indicators for observation and evaluation, such as the temperature of the exhaust gas outlet of the reduction furnace. This data represents the stability inside the reduction furnace. When the reduction furnace experiences some disturbances or abnormal temperature rises, the central control room can make a judgment based on some key data collected from the digital control system (DCS).

[0003] Human judgment has a lag effect. When an abnormality is detected, atomization occurs in the furnace. At this time, low-quality polysilicon has already been deposited on the silicon rod. Although normal production can be restored by timely adjustment, some low-quality polysilicon has already been deposited. They will seriously affect the subsequent silicon participation in the reaction, ultimately leading to a significant drop in the quality of this batch of polysilicon. Summary of the Invention

[0004] The purpose of this invention is to provide a method and apparatus for predicting atomization in a reduction furnace, so as to solve the problem that current technology cannot accurately predict atomization reactions in a reduction furnace, and to improve the problem that atomization conditions are difficult to quantify.

[0005] To achieve the above objectives, embodiments of the present invention provide a method for predicting atomization in a reduction furnace, comprising:

[0006] Based on the prediction model, the theoretical discharge temperature of the reduction exhaust gas at the first moment is determined;

[0007] The predicted temperature for the second time is determined based on the actual exhaust temperature of the reducing exhaust gas at the first time and the theoretical exhaust temperature of the reducing exhaust gas at the first time; the second time is the time after the first time.

[0008] Based on the predicted temperature, the atomization reaction of the reduction furnace at the second time is predicted.

[0009] Optionally, the method further includes: constructing a prediction model, wherein constructing the prediction model includes:

[0010] A prediction model is constructed based on the assumption that the input heat equals the output heat within a preset time period. The preset time period is the period from the point where the total electrical power in the reduction process passes through the inflection point from rising to falling to the furnace shutdown. The input heat includes the heat dissipated by heating the silicon core through the current. The output heat includes the heat obtained from the furnace drum water, the heat obtained from the chassis water, the heat discharged from the reduction tail gas, and the heat of reaction.

[0011] Optionally, the input heat is determined based on the input current value, the real-time resistance of the conductor, and the reaction time; the real-time resistance of the conductor is the resistance of the silicon core and the polycrystalline silicon accumulated through the vapor deposition reaction.

[0012] The process of obtaining the real-time resistance of a conductor includes:

[0013] Empirical parameters are obtained; these empirical parameters are derived by reverse calculation from the input total power curve; the total power curve is determined by the material list and obtained through multiple preset experiments.

[0014] Obtain the length, cross-sectional area, and conductivity of the silicon core;

[0015] The ratio between the first product of the empirical parameters and the length of the silicon core, and the second product of the cross-sectional area and the conductivity of the silicon core, is determined as the real-time resistance of the conductor.

[0016] Optionally, the heat discharged from the reduction tail gas is determined based on the product of the amount of substance, the gas constant, and the temperature difference, wherein the amount of substance and the gas constant are constants, and the temperature difference is the difference between the furnace temperature and the tail gas temperature.

[0017] Optionally, determining the predicted temperature for the second time period based on the actual discharge temperature of the reducing exhaust gas at the first time point and the theoretical discharge temperature of the reducing exhaust gas includes:

[0018] Determine the first difference between the actual discharge temperature of the reduction exhaust gas and the theoretical discharge temperature of the reduction exhaust gas;

[0019] The deviation equation is determined by fitting a preset function based on the first difference and the total power curve; the total power curve is determined by the material list and obtained through multiple preset experiments.

[0020] Based on the deviation equation and the current exhaust gas temperature, the predicted temperature for the second time period is determined.

[0021] Optionally, predicting the atomization reaction in the reduction furnace at the second time based on the predicted temperature includes:

[0022] The reaction temperature of the first feed and the content ratio of the second feed that undergo atomization reaction in the reduction furnace are determined; the reaction temperature of the second feed is greater than that of the first feed; both the first feed and the second feed will generate amorphous silicon when decomposed in the reduction furnace.

[0023] Based on the predicted temperature, the real-time temperature inside the reduction furnace is determined.

[0024] Based on the real-time temperature inside the furnace, the reaction temperature of the first feed, and the content ratio of the second feed, the atomization reaction of the reduction furnace at the second time is predicted.

[0025] Optionally, the reaction temperature of the first feed that undergoes atomization in the reduction furnace and the content ratio of the second feed are determined, including:

[0026] The concentration of amorphous silicon was determined using a dust detector in the exhaust gas duct.

[0027] Based on the concentration of the amorphous silicon, a first preset algorithm is used to calibrate and determine the reaction temperature of the first feed and the content ratio of the second feed that undergoes atomization reaction in the reduction furnace.

[0028] Optionally, the in-furnace reaction during the first preset time after furnace start-up is unrelated to the atomization reaction, while the in-furnace reaction during the second preset time after furnace start-up is related to the atomization reaction, wherein the second preset time is the time after the first preset time.

[0029] To achieve the above objectives, embodiments of the present invention also provide a predictive device for atomization in a reduction furnace, comprising:

[0030] The first determining module is used to determine the theoretical discharge temperature of the reduced exhaust gas at the first time based on the prediction model.

[0031] The second determining module is used to determine the predicted temperature for the second time based on the actual discharge temperature of the reducing exhaust gas at the first time and the theoretical discharge temperature of the reducing exhaust gas at the first time; the second time is the time after the first time.

[0032] The first processing module is used to predict the atomization reaction of the reduction furnace at the second time based on the predicted temperature.

[0033] To achieve the above objectives, embodiments of the present invention also provide a readable storage medium having a program or instructions stored thereon, which, when executed by a processor, implement the steps in the prediction method for atomization of a reduction furnace as described in any of the preceding claims.

[0034] The beneficial effects of the above-described technical solution of the present invention are as follows:

[0035] In this embodiment of the invention, the theoretical discharge temperature of the reduction tail gas at the first time is determined according to the prediction model; the predicted temperature at the second time is determined based on the actual discharge temperature and the theoretical discharge temperature of the reduction tail gas at the first time; the second time is the time after the first time; the atomization reaction of the reduction furnace at the second time is predicted based on the predicted temperature. Here, the predicted temperature is used to accurately determine the atomization degree of the reduction furnace, so that the operator can grasp the atomization degree in the reduction furnace in time and make adjustments, so as to achieve a certain degree of atomization boundary control operation, so as to reduce power consumption and ensure the quality of the reduction furnace products. Attached Figure Description

[0036] Figure 1 A schematic flowchart of the method for predicting atomization in a reduction furnace provided in an embodiment of the present invention;

[0037] Figure 2 This is a schematic diagram of the structure of the prediction device for a reduction furnace atomization provided in an embodiment of the present invention. Detailed Implementation

[0038] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0039] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the invention. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification do not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0040] In various embodiments of the present invention, it should be understood that the sequence number of each process described below does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0041] like Figure 1 As shown, an embodiment of the present invention provides a method for predicting atomization in a reduction furnace, comprising:

[0042] Step 11: Determine the theoretical exhaust temperature of the reducing exhaust gas at the first moment based on the prediction model.

[0043] Step 12: Determine the predicted temperature for the second time based on the actual exhaust temperature of the reduced exhaust gas at the first time and the theoretical exhaust temperature of the reduced exhaust gas; the second time is the time after the first time.

[0044] Step 13: Based on the predicted temperature, predict the atomization reaction of the reduction furnace at the second time.

[0045] In this embodiment of the invention, the first time can be understood as the current moment, and the second time can be understood as a certain period of time in the future. By using the actual value and the theoretical value at the current moment, a deviation equation between the actual and theoretical values ​​is determined. Based on this deviation equation, a predicted temperature for a certain future time is preset. Since there is a corresponding relationship between the predicted temperature and the atomization reaction, the predicted temperature can be used to accurately predict whether an atomization reaction will occur in the reduction furnace at a certain future time, and accurately determine the degree of atomization in the reduction furnace. This allows operators to promptly grasp the degree of atomization in the reduction furnace and make adjustments, achieving a certain degree of atomization boundary control operation, thereby reducing power consumption while ensuring the quality of the reduction furnace products.

[0046] Optionally, the above method further includes: constructing a prediction model, wherein constructing the prediction model includes:

[0047] A prediction model is constructed based on the assumption that the input heat equals the output heat within a preset time period. The preset time period is the period from the point where the total electrical power in the reduction process passes through the inflection point from rising to falling to the furnace shutdown. The input heat includes the heat dissipated by heating the silicon core through the current. The output heat includes the heat obtained from the furnace drum water, the heat obtained from the chassis water, the heat discharged from the reduction tail gas, and the heat of reaction.

[0048] Here, the aforementioned preset time period refers to the time period from when the total power output passes the inflection point from rising to falling until the furnace is shut down.

[0049] In this embodiment of the invention, a prediction model is constructed based on the premise that the input heat is equal to the output heat within a preset time period. That is, it is applicable to the time period from the inflection point of the total power from rising to falling to the furnace shutdown. According to the law of conservation of energy, the input heat is equal to the output heat, and a prediction model is constructed. The input heat mainly includes the heat dissipated by heating the silicon core through the current, and the output heat mainly includes the heat obtained by the furnace water, the heat obtained by the chassis water, the heat discharged by the reduction tail gas (i.e. the heat carried out by the discharged reduction tail gas), and the heat of reaction.

[0050] It should be noted that the main functions of the furnace water and the bottom water are to ensure that the reaction temperature inside the furnace remains stable and to remove excess heat from the reduction furnace through heat exchange. The temperatures of the furnace water and the bottom water are calibrated by temperature monitoring instruments at the input and output ends, and the mass is obtained by the cross-sectional area and flow rate of the upper and lower water pipes. Since a certain amount of additives are added to the water, the specific heat capacity of the furnace water and the bottom water is obtained experimentally. In summary, the heat carried out by the furnace water and the bottom water can be regarded as a known quantity.

[0051] It should also be noted that the heat of reaction can be calculated according to theoretical formulas. The forward reaction conversion rate can be calculated from parameters such as the weight of silicon rods after the furnace is shut down, and is considered a known quantity. Under the condition that the material table remains unchanged, the forward reaction conversion rate of each furnace will have slight fluctuations, but the changes are not significant and have little impact.

[0052] Optionally, the input heat is determined based on the input current value, the real-time resistance of the conductor, and the reaction time; the real-time resistance of the conductor is the resistance of the silicon core and the polycrystalline silicon accumulated through the vapor deposition reaction.

[0053] The process of obtaining the real-time resistance of a conductor includes:

[0054] Empirical parameters are obtained; these empirical parameters are derived by reverse calculation from the input total power curve; the total power curve is determined by the material list and obtained through multiple preset experiments.

[0055] Obtain the length, cross-sectional area, and conductivity of the silicon core;

[0056] The ratio between the first product of the empirical parameters and the length of the silicon core, and the second product of the cross-sectional area and the conductivity of the silicon core, is determined as the real-time resistance of the conductor.

[0057] In this embodiment of the invention, the input heat is mainly caused by the power supply. During this process, the input heat is determined based on the current value at the input terminal, the real-time resistance of the conductor, and the reaction time. That is, input heat = current squared × conductor real-time resistance × time. The magnitude of the current is known; an ammeter is added at the input terminal, and the reading of the ammeter is the magnitude of the current. The conductor real-time resistance mainly refers to the resistance of the silicon core and the polycrystalline silicon accumulated through the vapor deposition reaction. Within 1480℃, the conductivity of silicon can be calculated from parameters such as electron mobility, hole mobility, and intrinsic carrier concentration. The relevant parameters can be theoretically calculated by referring to tables. Because the reaction temperature of each reaction stage is constant (<1480℃), the conductivity of silicon fluctuates within a small range and has little impact on the resistance value.

[0058] It is important to note that due to the influence of factors such as the gas field and temperature field on the vapor deposition rate, and the complexity of the reaction mechanism, the vapor deposition reaction rate is almost impossible to calibrate in real time. Therefore, it is necessary to add empirical parameters based on historical data. The real-time conductor resistance = empirical parameter × silicon core length / (silicon core cross-sectional area × silicon core conductivity). The empirical parameter in the input heat is derived by back-calculating the total input electrical power. The total electrical power curve is determined by the material table and obtained through multiple actual experiments.

[0059] Optionally, the heat discharged from the reduction tail gas is determined based on the product of the amount of substance, the gas constant, and the temperature difference, wherein the amount of substance and the gas constant are constants, and the temperature difference is the difference between the furnace temperature and the tail gas temperature.

[0060] In this embodiment, the process of reducing tail gas discharge can be regarded as the discharge of an isobaric ideal gas. According to the ideal gas law, the heat that the reducing tail gas can carry out = amount of substance × gas constant × (furnace temperature - tail gas temperature). Here, the amount of substance and the gas constant are both constants. The furnace temperature fluctuates within a small range, so the temperature of the reducing tail gas discharge can be obtained.

[0061] Optionally, step 12 above includes:

[0062] Determine the first difference between the actual discharge temperature of the reduction exhaust gas and the theoretical discharge temperature of the reduction exhaust gas;

[0063] The deviation equation is determined by fitting a preset function based on the first difference and the total power curve; the total power curve is determined by the material list and obtained through multiple preset experiments; here, the total power curve is the input total power curve.

[0064] Based on the deviation equation and the current exhaust gas temperature, the predicted temperature for the second time period is determined.

[0065] In this embodiment of the invention, the theoretical exhaust temperature of the reduced exhaust gas is determined according to the prediction model. That is, the exhaust temperature of the reduced exhaust gas is a theoretical value. Since the total power curve is corrected based on experiments, a temperature monitoring instrument can be set at the exhaust gas discharge location. The difference between the actual value and the theoretical value is determined as the first difference. Based on the first difference and the total power curve, a preset function is fitted to determine the deviation equation. This can be understood as follows: based on the first difference and correlated with the total power curve, after accumulating enough data, the data is fitted with a polynomial function to determine the deviation equation. Based on the deviation equation and the current exhaust gas temperature, the predicted temperature at the second time can be predicted, thereby obtaining the exhaust gas temperature at the second time.

[0066] In one optional embodiment, the prediction model in this embodiment of the invention cleans the input data to ensure the reliability of the model output.

[0067] Specifically, during the data collection process, the sensor readings become unstable due to excessive deviation caused by nitrogen-hydrogen replacement during furnace start-up, leading to abnormal data fluctuations and data distortion. Furthermore, the cumulative effect before and after furnace start-up and shutdown is not significant, and the data is very close to the theoretical value. Therefore, the data before and after furnace start-up and shutdown can be directly replaced by the theoretical value. During operation, the LOF (Local Outlier Factor) algorithm (i.e., the local outlier factor detection method) is used to detect anomalies. By discarding data with local density anomalies, reliable data is obtained. For data gaps, multiple sets of data are used for cleaning to obtain complete data segments.

[0068] In one optional embodiment, the in-furnace reaction during a first preset time after furnace start-up is unrelated to the atomization reaction, while the in-furnace reaction during a second preset time after furnace start-up is related to the atomization reaction, wherein the second preset time is the time after the first preset time.

[0069] Here, the first preset time is the first 30 hours after the furnace is started. During the reaction process, the phenomenon that needs to be paid attention to in the first 30 hours after the furnace is started is not atomization. The second preset time is after 30 hours after the furnace is started. The temperature of the reduction tail gas is only related to the atomization phenomenon after 30 hours after the furnace is started.

[0070] Due to the reduction process, the temperature of the reduction tail gas can reflect part of the atomization situation inside the furnace. When atomization occurs, the tail gas temperature rises and the furnace water temperature drops. Therefore, predicting the reduction tail gas temperature can provide an early warning of atomization phenomena in the future, giving operators sufficient reaction time to adjust parameters to avoid atomization. Optionally, step 13 above includes:

[0071] The reaction temperature of the first feed and the content ratio of the second feed that undergo atomization reaction in the reduction furnace are determined; the reaction temperature of the second feed is greater than that of the first feed; both the first feed and the second feed will generate amorphous silicon when decomposed in the reduction furnace.

[0072] Based on the predicted temperature, the real-time temperature inside the reduction furnace is determined.

[0073] Based on the real-time temperature inside the furnace, the reaction temperature of the first feed, and the content ratio of the second feed, the atomization reaction of the reduction furnace at the second time is predicted.

[0074] In this embodiment of the invention, the first feed can be trichlorosilane (TCS), and the second feed can be dichlorosilane (DCS). During the process of trichlorosilane (TCS) entering the reduction furnace, it often contains a certain proportion of dichlorosilane (DCS). During the operation of the reduction furnace, excessively low temperatures will cause DCS to decompose preferentially, generating amorphous silicon. That is, under the same reaction conditions, dichlorosilane (DCS) is more reactive and more prone to bulk reactions, leading to atomization. In the later stages of the reaction, as the temperature gradually decreases, excess TCS will decompose into amorphous silicon. The phenomenon of amorphous silicon wandering within the furnace is called atomization. Here, based on the predicted temperature, the real-time temperature inside the reduction furnace is determined. For example, if the predicted temperature for a future period is T1, the real-time temperature inside the reduction furnace corresponding to T1 can be determined, i.e., the current decomposition temperature of the reduction furnace is determined. Based on the real-time temperature inside the furnace, the reaction temperature of the first feed, and the content ratio of the second feed, the atomization reaction in the reduction furnace at the second time is predicted. For example, if the real-time temperature inside the furnace is determined to be 850-950 degrees Celsius, and the content of dichlorosilane (DCS) is determined to be controlled within 4% using logistic regression analysis, and the content ratio of the second feed is also determined to be within 4%, if it exceeds 4%, then an atomization reaction is predicted to occur in the reduction furnace at the second time, and corresponding adjustment strategies need to be made in a timely manner.

[0075] Optionally, the reaction temperature of the first feed that undergoes atomization in the reduction furnace and the content ratio of the second feed are determined, including:

[0076] The concentration of amorphous silicon was determined using a dust detector in the exhaust gas duct.

[0077] Based on the concentration of the amorphous silicon, a first preset algorithm is used to calibrate and determine the reaction temperature of the first feed and the content ratio of the second feed that undergoes atomization reaction in the reduction furnace.

[0078] It should be noted that in this invention, by installing a dust detector on the polysilicon reduction furnace, the concentration of dust particles (fine silicon powder) inside the furnace can theoretically be accurately detected. This allows operators to adjust the furnace promptly, achieving a certain degree of edge-controlled operation at the furnace's atomization boundary, thus reducing power consumption and ensuring product quality. However, existing dust detectors have limited high-temperature resistance and cannot be directly installed on the furnace body. Therefore, the atomization situation inside the furnace is determined by measuring the concentration of amorphous silicon using a dust detector. Here, multiple batches of dust particle concentration data monitored from the exhaust gas outlet pipeline of the polysilicon reduction furnace can be analyzed and studied to establish the relationship between the concentrations of multiple dust particles monitored from the exhaust gas outlet pipeline and the atomization of the polysilicon reduction furnace.

[0079] In this embodiment of the invention, amorphous silicon is mixed in with the reduction tail gas and discharged into the furnace. By installing a particulate dust detector on the reduction tail gas pipeline, the atomization situation in the furnace is determined by detecting the concentration of amorphous silicon. The variables that affect the atomization situation in the furnace are the temperature that leads to the decomposition of excessive TCS and the content of dichlorosilane (DCS). For ordered multi-category variables, the Logistic regression method is used for calibration to solve the control range of the temperature for the decomposition of excessive TCS and the content of dichlorosilane (DCS).

[0080] The composition of the feed material is analyzed periodically upon entry, and the content of dichlorosilane (DCS) in the feed can be considered known. Therefore, controlling the DCS content in the feed material can eliminate the influence of this variable. By predicting the temperature of the reduction tail gas over a certain period of time, the temperature change inside the reduction furnace over a certain period of time can be inferred. This temperature change is then compared with the temperature value of excess TCS decomposition, thereby guiding the internal operators to perform relevant operations.

[0081] In one specific implementation, the phenomenon that needs attention in the first 30 hours after the furnace is started during the reaction process is not atomization. Therefore, the temperature of the reduction tail gas is only linked to the atomization phenomenon after 30 hours of furnace start-up. The present invention provides a material table, as shown in Table 1 below. Table 1 shows the relevant instrument measurement values ​​based on a certain material table for a certain period of time (4-minute interval).

[0082] Table 1:

[0083]

[0084] Among them, A1 to C1 and A2 to C2 are multiple electrodes of the reduction furnace, which are conventional techniques in this field and will not be explained in detail here.

[0085] The cross-sectional area of ​​the silicon rod can be considered to increase linearly after 30 hours of furnace start-up. The initial diameter is about 3 cm, and the final diameter after 100 hours is about 15 cm. The length is a fixed value of 225 cm. The empirical parameter is fitted as the product of a fixed coefficient K and time t. Substituting the above values ​​into the formula, the theoretical value of the reduction tail gas at the current moment is obtained. Compared with the actual value, the deviation equation is obtained: e = -0.625Kt + 5. From this, the value of the reduction tail gas in the future period (T + e) ​​can be estimated, where e is the deviation and T is the temperature of the reduction tail gas at the current moment.

[0086] Based on logistic regression analysis, the temperature limit for excessive TCS decomposition is determined to be 850-950 degrees Celsius, and the dichlorosilane (DCS) content should be controlled below 4%. Using the predicted temperature as a working model, the real-time furnace temperature can be calculated. When the real-time furnace temperature is below 950 degrees Celsius, the operator needs to constantly monitor the TCS feed rate to prevent atomization caused by excessive TCS self-decomposition. By reminding the operator to constantly monitor the TCS feed rate for a period of time, sufficient reaction and operation time is allowed.

[0087] In summary, the method of the present invention can accurately determine the atomization degree of the reduction furnace, so that operators can promptly grasp the atomization degree inside the reduction furnace and make timely adjustments, achieving a certain degree of atomization boundary control operation, thereby reducing power consumption and ensuring the quality of products from the reduction furnace. This method greatly improves the control quality of the reduction furnace and reduces the difficulty of operating the reduction furnace.

[0088] Reference Figure 2 As shown, this embodiment of the invention also provides a predictive device for atomization in a reduction furnace, comprising:

[0089] The first determining module 21 is used to determine the theoretical discharge temperature of the reduction exhaust gas at the first time according to the prediction model.

[0090] The second determining module 22 is used to determine the predicted temperature for the second time based on the actual discharge temperature of the reducing exhaust gas at the first time and the theoretical discharge temperature of the reducing exhaust gas at the first time; the second time is the time after the first time.

[0091] The first processing module 23 is used to predict the atomization reaction of the reduction furnace at the second time based on the predicted temperature.

[0092] Optionally, the aforementioned predictive device for atomization in the reduction furnace further includes:

[0093] The prediction module is used to build prediction models, which includes:

[0094] A prediction model is constructed based on the assumption that the input heat equals the output heat within a preset time period. The preset time period is the period from the point where the total electrical power in the reduction process passes through the inflection point from rising to falling to the furnace shutdown. The input heat includes the heat dissipated by heating the silicon core through the current. The output heat includes the heat obtained from the furnace drum water, the heat obtained from the chassis water, the heat discharged from the reduction tail gas, and the heat of reaction.

[0095] Optionally, the input heat is determined based on the input current value, the real-time resistance of the conductor, and the reaction time; the real-time resistance of the conductor is the resistance of the silicon core and the polycrystalline silicon accumulated through the vapor deposition reaction.

[0096] The aforementioned prediction module includes:

[0097] The first acquisition unit is used to acquire empirical parameters; the empirical parameters are derived by reverse calculation from the input total power curve; the total power curve is determined by the material list and obtained through multiple preset experiments;

[0098] The second acquisition unit is used to acquire the length, cross-sectional area, and conductivity of the silicon core.

[0099] A first processing unit is configured to determine the real-time resistance of the conductor by a ratio between a first product of the empirical parameters and the length of the silicon core, and a second product of the cross-sectional area of ​​the silicon core and the conductivity of the silicon core.

[0100] Optionally, the heat discharged from the reduction tail gas is determined based on the product of the amount of substance, the gas constant, and the temperature difference, wherein the amount of substance and the gas constant are constants, and the temperature difference is the difference between the furnace temperature and the tail gas temperature.

[0101] Optionally, the second determining module 22 described above includes:

[0102] The first determining unit is used to determine a first difference between the actual discharge temperature of the reducing exhaust gas and the theoretical discharge temperature of the reducing exhaust gas.

[0103] The second determining unit is used to perform a preset function fitting based on the first difference and the total power curve to determine the deviation equation; the total power curve is determined by the material list and obtained through multiple preset experiments;

[0104] The third determining unit is used to determine the predicted temperature at the second time based on the deviation equation and the current exhaust gas temperature.

[0105] Optionally, the first processing module 23 described above includes:

[0106] The fourth determining unit is used to determine the reaction temperature of the first feed and the content ratio of the second feed that undergo atomization reaction in the reduction furnace; the reaction temperature of the second feed is greater than the reaction temperature of the first feed; both the first feed and the second feed will generate amorphous silicon when decomposed in the reduction furnace.

[0107] The fifth determining unit is used to determine the real-time temperature inside the reduction furnace based on the predicted temperature.

[0108] The second processing unit is used to predict the atomization reaction of the reduction furnace at the second time based on the real-time temperature inside the furnace, the reaction temperature of the first feed, and the content ratio of the second feed.

[0109] Optionally, the fourth determining unit described above includes:

[0110] The first determining subunit is used to determine the concentration of amorphous silicon based on the dust detector in the reduction exhaust gas pipeline;

[0111] The second determining subunit is used to determine the reaction temperature of the first feed and the content ratio of the second feed that undergo atomization reaction in the reduction furnace based on the concentration of the amorphous silicon and by calibration using a first preset algorithm.

[0112] Optionally, the in-furnace reaction during the first preset time after furnace start-up is unrelated to the atomization reaction, while the in-furnace reaction during the second preset time after furnace start-up is related to the atomization reaction, wherein the second preset time is the time after the first preset time.

[0113] The implementation embodiments of the above-mentioned reduction furnace atomization prediction method are all applicable to the embodiments of the reduction furnace atomization prediction device, and can achieve the same technical effect.

[0114] An embodiment of the present invention provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the steps in the prediction method for atomization of the reduction furnace as described above and achieve the same technical effect. To avoid repetition, the details will not be repeated here.

[0115] The processor mentioned above is the processor in the reduction furnace atomization prediction method described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0116] In this embodiment of the invention, the module can be implemented in software so that it can be executed by various types of processors. For example, an identified executable code module may include one or more physical or logical blocks of computer instructions, which may be constructed as objects, procedures, or functions. Nevertheless, the executable code of the identified module does not need to be physically located together, but may include different instructions stored in different bits, which, when logically combined, constitute the module and achieve the module's intended purpose.

[0117] In practice, an executable code module can be a single instruction or many instructions, and can even be distributed across multiple different code segments, different programs, and across multiple memory devices. Similarly, operational data can be identified within the module and can be implemented in any suitable form and organized within any suitable data structure. This operational data can be collected as a single dataset or distributed across different locations (including different storage devices), and can exist, at least in part, solely as electronic signals within the system or network.

[0118] When a module can be implemented using software, considering the current level of hardware technology, modules that can be implemented in software can be implemented using hardware circuits by those skilled in the art to achieve the corresponding functions, without considering cost. These hardware circuits include conventional very-large-scale integrated circuits (VLSI) or gate arrays, as well as existing semiconductors such as logic chips and transistors, or other discrete components. Modules can also be implemented using programmable hardware devices, such as field-programmable gate arrays, programmable array logic, and programmable logic devices.

[0119] The exemplary embodiments described above are with reference to the accompanying drawings. Many different forms and embodiments are feasible without departing from the spirit and teachings of the invention. Therefore, the invention should not be construed as limiting the exemplary embodiments set forth herein. Rather, these exemplary embodiments are provided to make the invention complete and convey the scope of the invention to those skilled in the art. In these drawings, component dimensions and relative dimensions may be exaggerated for clarity. The terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to be limiting. As used herein, unless clearly indicated otherwise, the singular forms “a,” “an,” and “the” are intended to include all such forms. It will be further understood that the terms “comprising” and / or “including”, when used in this specification, indicate the presence of the stated features, integers, steps, operations, components, and / or elements, but do not exclude the presence or addition of one or more other features, integers, steps, operations, components, and / or groups thereof. Unless otherwise indicated, when stated, a range of values ​​includes the upper and lower limits of the range and any subranges in between.

[0120] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for predicting atomization in a reduction furnace, characterized in that, include: Based on the prediction model, the theoretical discharge temperature of the reduction tail gas at the first moment is determined. The prediction model is constructed by: assuming that the input heat equals the output heat within a preset time period; wherein the preset time period is the period from the inflection point of the total electrical power in the reduction process from rising to falling to the furnace shutdown; the input heat includes the heat dissipated by heating the silicon core through the current; the output heat includes the heat obtained from the furnace drum water, the heat obtained from the chassis water, the heat discharged from the reduction tail gas, and the heat of reaction; the heat discharged from the reduction tail gas is determined based on the product of the amount of substance, the gas constant, and the temperature difference, wherein the amount of substance and the gas constant are constants, and the temperature difference is the difference between the furnace temperature and the tail gas temperature. Determining the predicted temperature for the second time based on the actual discharge temperature of the reduction exhaust gas at the first time and the theoretical discharge temperature of the reduction exhaust gas at the first time includes: determining a first difference between the actual discharge temperature of the reduction exhaust gas and the theoretical discharge temperature of the reduction exhaust gas; performing a preset function fitting based on the first difference and the total power curve to determine a deviation equation; the total power curve is determined by the material table and obtained through multiple preset experiments; determining the predicted temperature for the second time based on the deviation equation and the current temperature of the reduction exhaust gas; the second time is the time after the first time. Based on the predicted temperature, predicting the atomization reaction in the reduction furnace at the second time includes: determining the reaction temperature of the first feed entering the reduction furnace and the content ratio of the second feed; the reaction temperature of the second feed is greater than the reaction temperature of the first feed; both the first feed and the second feed decompose in the reduction furnace to generate amorphous silicon; determining the real-time temperature inside the reduction furnace based on the predicted temperature; and predicting the atomization reaction in the reduction furnace at the second time based on the real-time temperature inside the furnace, the reaction temperature of the first feed, and the content ratio of the second feed.

2. The method according to claim 1, characterized in that, The input heat is determined based on the input current value, the real-time resistance of the conductor, and the reaction time; the real-time resistance of the conductor is the resistance of the silicon core and the polycrystalline silicon accumulated through the vapor deposition reaction. The process of obtaining the real-time resistance of a conductor includes: Empirical parameters are obtained; these empirical parameters are derived by reverse calculation from the input total power curve; the total power curve is determined by the material list and obtained through multiple preset experiments. Obtain the length, cross-sectional area, and conductivity of the silicon core; The ratio between the first product of the empirical parameters and the length of the silicon core, and the second product of the cross-sectional area and the conductivity of the silicon core, is determined as the real-time resistance of the conductor.

3. The method according to claim 1, characterized in that, Determine the reaction temperature of the first feed that undergoes atomization in the reduction furnace and the content ratio of the second feed, including: The concentration of amorphous silicon was determined using a dust detector in the exhaust gas duct. Based on the concentration of the amorphous silicon, a first preset algorithm is used to calibrate and determine the reaction temperature of the first feed and the content ratio of the second feed that undergoes atomization reaction in the reduction furnace.

4. The method according to claim 1, characterized in that, The furnace reaction within the first preset time after furnace start-up is unrelated to the atomization reaction, while the furnace reaction within the second preset time after furnace start-up is related to the atomization reaction. The second preset time is the time after the first preset time.

5. A predictive device for atomization in a reduction furnace, characterized in that, include: The first determining module is used to determine the theoretical discharge temperature of the reduced exhaust gas at the first time based on the prediction model. The second determining module is used to determine the predicted temperature for the second time based on the actual discharge temperature of the reducing exhaust gas at the first time and the theoretical discharge temperature of the reducing exhaust gas. The second time is the time after the first time; The second determining module includes: a first determining unit, used to determine a first difference between the actual discharge temperature of the reduction tail gas and the theoretical discharge temperature of the reduction tail gas; a second determining unit, used to perform a preset function fitting based on the first difference and the total power curve to determine a deviation equation; the total power curve is determined by the material table and obtained through multiple preset experiments; and a third determining unit, used to determine the predicted temperature at a second time based on the deviation equation and the current temperature of the reduction tail gas. A first processing module is used to predict the atomization reaction of the reduction furnace at the second time based on the predicted temperature. The first processing module includes: a fourth determining unit, used to determine the reaction temperature of the first feed entering the reduction furnace and the content ratio of the second feed; the reaction temperature of the second feed is greater than the reaction temperature of the first feed; both the first feed and the second feed decompose in the reduction furnace to generate amorphous silicon; a fifth determining unit, used to determine the real-time temperature inside the reduction furnace based on the predicted temperature; and a second processing unit, used to predict the atomization reaction of the reduction furnace at the second time based on the real-time temperature inside the furnace, the reaction temperature of the first feed, and the content ratio of the second feed. The prediction device for atomization in the reduction furnace further includes a prediction module for constructing a prediction model. The prediction model is constructed based on the assumption that the input heat equals the output heat within a preset time period. The preset time period is the period from the point where the total electrical power in the reduction process reaches its inflection point (from rising to falling) to the furnace shutdown. The input heat includes the heat dissipated by heating the silicon core through current. The output heat includes the heat obtained from the furnace drum water, the heat obtained from the chassis water, the heat discharged from the reduction tail gas, and the heat of reaction. The heat discharged from the reduction tail gas is determined based on the product of the amount of substance, the gas constant, and the temperature difference, where the amount of substance and the gas constant are constants, and the temperature difference is the difference between the furnace temperature and the tail gas temperature.

6. A readable storage medium having a program or instructions stored thereon, characterized in that, When the program or instructions are executed by the processor, they implement the steps in the method for predicting atomization of a reduction furnace as described in any one of claims 1 to 4.

Images