Online monitoring and feedback control method for plastic bottle flake cleaning equipment

Abstract

The present application relates to the technical field of chemical control, in particular to an online monitoring and feedback control method of plastic bottle piece cleaning equipment, which divides continuous cleaning stages based on statistical distribution jump of heating power and stirring current variation, effectively capturing the dynamic changes caused by material entering; by analyzing the consumption continuation relationship and time sequence evolution difference between adjacent stages, respectively determining the reaction residual degree and consumption trend strength, accurately quantifying the actual consumption change and residual influence within the low-frequency sampling interval; then combining with the accumulated concentration of invalid by-products to comprehensively determine the alkali supplement tendency coefficient, and based on the coefficient and the free alkali concentration to determine the target alkali supplement amount. The method incorporates the real dynamic consumption rhythm and by-product accumulation constraints into the control link, making the accuracy of the automatic alkali supplement control of the present application higher.

Description

Technical Field

[0001] This invention relates to the field of chemical control technology, specifically to an online monitoring and feedback control method for plastic bottle flake washing equipment. Background Technology

[0002] In the cleaning and recycling process of plastic bottle flakes, alkaline solutions are typically used to remove oil and adhesives from the surface of the flakes, and the composition of the solution must be monitored in real time to maintain the cleaning activity of the equipment. Existing technologies usually employ an automatic alkali replenishment control method based on low-frequency sampling concentration deviation. Specifically, this relies on an online analyzer to detect the free alkali concentration of the solution at a fixed sampling period. When the detected concentration value is lower than the set target, the required amount of alkali to be replenished is calculated unidirectionally based on the concentration deviation and the dosing action is activated.

[0003] However, existing technologies do not take into account the phased dynamic fluctuations in the cleaning process as materials enter in batches. They rely solely on a single low-frequency concentration deviation, ignoring the changes in the actual consumption rate of free alkali due to material impact within the low-frequency sampling interval, as well as the lingering effects of incompletely reacted residues from the previous batch. Furthermore, they do not consider the drastic compression of the accumulation space for ineffective byproducts during the alkali replenishment process. This makes it impossible for the alkali replenishment action to proactively match the actual material consumption rhythm, which can easily lead to excessive alkali replenishment and the blind and accelerated accumulation of ineffective byproducts. Consequently, the accuracy of existing technologies for automatic alkali replenishment control based on low-frequency sampling concentration deviations is relatively low. Summary of the Invention

[0004] To address the low accuracy of existing technologies that rely on low-frequency sampling concentration deviation for automatic alkali replenishment control, this invention aims to provide an online monitoring and feedback control method for plastic bottle flake washing equipment. The specific technical solution adopted is as follows: This invention provides an online monitoring and feedback control method for a plastic bottle flake washing equipment, comprising: During the operation of the plastic bottle flake washing equipment, the free alkali concentration and cumulative concentration of ineffective byproducts of the washing equipment at each low-frequency sampling time, as well as the heating power variation and stirring current variation at each high-frequency sampling time are obtained. Based on the statistical distribution jumps of heating power variation and stirring current variation over time, all continuous cleaning stages are divided; based on the consumption continuity relationship between each continuous cleaning stage and the previous continuous cleaning stage in terms of stirring current variation and free alkali concentration, the corresponding reaction residue degree is determined. Based on the temporal evolution difference of the stirring current variation between each continuous cleaning stage and the previous continuous cleaning stage, the corresponding consumption trend intensity is determined; based on the reaction residue level, the consumption trend intensity, and the cumulative concentration of ineffective byproducts, the corresponding alkali replenishment tendency coefficient is determined. Based on the alkali replenishment tendency coefficient and the free alkali concentration, the target alkali replenishment amount is determined; based on the target alkali replenishment amount, the plastic bottle flake washing equipment is automatically controlled to replenish alkali.

[0005] Furthermore, the process of obtaining the continuous cleaning stage includes: Within a preset historical window before each high-frequency sampling moment, the power response intensity is determined based on the mean of all heating power variations; the heating power fluctuation is determined based on the standard deviation of all heating power variations; the current drive intensity is determined based on the mean of all stirring current variations; and the stirring current fluctuation is determined based on the standard deviation of all stirring current variations. The characteristic rise time is obtained; the values ​​corresponding to the characteristic rise time in terms of power response intensity, heating power fluctuation, current drive intensity and stirring current fluctuation are all greater than the previous high-frequency sampling time. Obtain the stage segmentation time; the preceding high-frequency sampling time of the stage segmentation time is not the feature rise time, and the subsequent consecutive preset number of high-frequency sampling times are all feature rise times; The process is divided into all continuous cleaning stages, with the stage segmentation time as the stage interval.

[0006] Furthermore, the process for obtaining the degree of reaction residue includes: Each consecutive cleaning stage is sequentially designated as the target cleaning stage, and the preceding consecutive cleaning stage is designated as the reference cleaning stage; the time of segmentation between the target cleaning stage and the reference cleaning stage is designated as the target segmentation time. The total concentration decrease in the target cleaning stage is determined based on the difference between the free alkali concentration at the last low-frequency sampling moment of the reference cleaning stage and the free alkali concentration at the first low-frequency sampling moment of the target cleaning stage. The total concentration decrease is allocated based on the change in stirring current before and after the target segment time, and the free alkali boundary concentration at the target segment time is determined. The actual consumption of the target cleaning stage is determined based on the difference between the free alkali boundary concentration at the target segment time and the free alkali boundary concentration at the corresponding next stage segment time. Obtain the consumption conversion coefficient corresponding to the target cleaning stage; determine the expected consumption based on the product of the consumption conversion coefficient and the cumulative value of the peak values ​​of all stirring current variations within the target cleaning stage; determine the degree of reaction residue in the target cleaning stage based on the relative deviation between the expected consumption and the actual consumption.

[0007] Furthermore, the process of obtaining the free alkali boundary concentration includes: The final impact intensity is determined by summing the peak values ​​of all stirring current variations between the last low-frequency sampling moment of the reference cleaning phase and the target segmentation moment; the first impact intensity is determined by summing the peak values ​​of all stirring current variations between the target segmentation moment and the first low-frequency sampling moment of the target cleaning phase; and the peak distribution weight at the target segmentation moment is determined based on the ratio between the final impact intensity and the first impact intensity. The boundary concentration consumption allocation at the target segment time is determined based on the product of the peak distribution ratio and the total concentration decrease; the free alkali boundary concentration at the target segment time is determined based on the difference between the free alkali concentration at the last low-frequency sampling time of the reference cleaning stage and the boundary concentration consumption allocation.

[0008] Furthermore, the process of obtaining the intensity of the consumption trend includes: Based on the number of peak values ​​of all stirring current variations that occur in each continuous cleaning stage and the total duration of the stage, the corresponding stage impact frequency is determined; based on the relative deviation between the stage impact frequency of each continuous cleaning stage and the stage impact frequency of the previous continuous cleaning stage, the corresponding frequency trend component is determined. The corresponding stage impact intensity is determined based on the average of the peak values ​​of all stirring current variations within each continuous cleaning stage; the corresponding intensity trend component is determined based on the relative deviation between the stage impact intensity of each continuous cleaning stage and the stage impact intensity of the previous continuous cleaning stage. The consumption trend intensity corresponding to each continuous cleaning stage is determined based on the sum of the frequency trend component and the intensity trend component.

[0009] Furthermore, the process of obtaining the alkali supplementation tendency coefficient includes: The first low-frequency sampling time of the next continuous cleaning stage of each continuous cleaning stage is used as the corresponding control reference time. Based on the relative deviation between the preset upper limit of the liquid replacement concentration and the cumulative concentration of ineffective byproducts at the control reference time, the corresponding liquid replacement constraint coefficient is determined. Based on the sum of the residual reaction level and the intensity of the consumption trend, a comprehensive compensation component for each continuous cleaning stage is determined; the product of the liquid exchange constraint coefficient and the comprehensive compensation component is positively correlated to determine the alkali replenishment tendency coefficient for each continuous cleaning stage.

[0010] Furthermore, the process of obtaining the target alkali supplementation amount includes: Calculate the difference between the preset target concentration of free alkali and the concentration of free alkali at the control reference time, and determine the first concentration deviation; Based on the first concentration deviation, the preset effective volume of the cleaning equipment, and the preset free alkali concentration of the alkali replenishing agent, the corresponding basic alkali replenishment amount is determined; wherein, the basic alkali replenishment amount is positively correlated with the first concentration deviation and the preset effective volume of the cleaning equipment, and negatively correlated with the preset free alkali concentration of the alkali replenishing agent; Based on the product of the basic alkali replenishment amount and the alkali replenishment tendency coefficient, the required alkali replenishment amount for each continuous cleaning stage is determined. Based on the preset upper limit of the liquid replacement concentration and the cumulative concentration of ineffective byproducts at the control reference time, the limit alkali replenishment volume for each continuous cleaning stage is determined. The target alkali replenishment amount for each continuous cleaning stage is determined based on the minimum value between the required alkali replenishment amount and the maximum alkali replenishment volume.

[0011] Furthermore, the process of obtaining the maximum alkali replenishment volume includes: Calculate the difference between the preset upper limit of the replacement solution concentration and the cumulative concentration of ineffective byproducts at the control reference time to determine the second concentration deviation; based on the second concentration deviation, the preset effective volume of the cleaning equipment, and the preset free alkali concentration of the alkali replenishing agent, determine the corresponding limit alkali replenishment volume; wherein, the limit alkali replenishment volume is positively correlated with the second concentration deviation and the preset effective volume of the cleaning equipment, and negatively correlated with the preset free alkali concentration of the alkali replenishing agent.

[0012] Furthermore, the process of obtaining the peak distribution weight includes: The total transition impact intensity is determined based on the sum of the final impact intensity and the initial impact intensity. The peak distribution ratio at the target segment time is determined based on the ratio between the final impact intensity and the sum of the transition impact intensities.

[0013] Furthermore, the process of obtaining the fluid exchange constraint coefficient includes: The corresponding capacity margin concentration is determined based on the difference between the preset upper limit of the fluid replacement concentration and the cumulative concentration of ineffective byproducts at the control reference time. The corresponding fluid replacement constraint coefficient is determined based on the ratio between the remaining capacity concentration and the preset upper limit of the fluid replacement concentration.

[0014] The present invention has the following beneficial effects: This invention divides the continuous cleaning stages based on the statistical distribution jumps of heating power and stirring current variations, effectively capturing the dynamic changes caused by material entry. By analyzing the consumption continuity and temporal evolution differences between adjacent stages, the degree of reaction residue and the intensity of consumption trend are determined respectively, accurately quantifying the actual consumption changes and residual effects within the low-frequency sampling interval. Furthermore, a supplementary alkali tendency coefficient is comprehensively determined by combining the cumulative concentration of ineffective byproducts, and the target supplementary alkali amount is determined based on this coefficient and the free alkali concentration. This method incorporates the real dynamic consumption rhythm and byproduct accumulation constraints into the control process, effectively avoiding excessive alkali supplementation and blind accumulation of byproducts caused by relying solely on low-frequency concentration deviations. It solves the technical problem of low accuracy in automatic alkali supplementation control based on low-frequency sampling concentration deviations in existing technologies, making the automatic alkali supplementation control of this application more accurate. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating an online monitoring and feedback control method for a plastic bottle flake washing device, as provided in one embodiment of the present invention. Detailed Implementation

[0016] The following description, in conjunction with the accompanying drawings, details the specific scheme of the online monitoring and feedback control method for a plastic bottle flake washing equipment provided by the present invention.

[0017] This invention provides an online monitoring and feedback control method for a plastic bottle flake washing equipment. Please refer to [link / reference]. Figure 1 The diagram illustrates a flowchart of an online monitoring and feedback control method for a plastic bottle flake washing device according to an embodiment of the present invention. The method includes: Step S101: During the operation of the plastic bottle flake cleaning equipment, obtain the free alkali concentration and cumulative concentration of ineffective byproducts at each low-frequency sampling time, as well as the heating power variation and stirring current variation at each high-frequency sampling time.

[0018] During the operation of plastic bottle flake cleaning equipment, real-time monitoring of the chemical state of the cleaning solution in the hot washing tank is required based on low-frequency cycles. Specifically, an online free alkali analyzer and an online total alkali analyzer are independently installed at the cleaning solution circulation loop or the sampling port on the side wall of the hot washing tank. During equipment operation, the online free alkali analyzer automatically extracts cleaning solution samples and performs titration testing according to a preset low-frequency sampling cycle to obtain the free alkali concentration at each low-frequency sampling time. Simultaneously, the online total alkali analyzer also tests the cleaning solution samples according to the same preset low-frequency sampling cycle to obtain the total alkali concentration at each low-frequency sampling time. To accurately quantify the newly generated contamination burden during the cleaning reaction, the total alkali concentration detected at each low-frequency sampling time is subtracted from the free alkali concentration to accurately determine the cumulative concentration of ineffective byproducts such as carbonates and saponifications continuously generated during the cleaning process at each low-frequency sampling time.

[0019] To compensate for the data lag caused by low-frequency chemical sampling analysis, and to simultaneously perform high-frequency dynamic monitoring of the electromechanical operating status of the cleaning equipment, continuously reflecting the instantaneous impact and load changes of materials within the cleaning tank, a power transmitter is connected to the temperature control unit of the hot washing tank heating system, and a current transmitter is connected to the frequency converter drive unit of the stirring motor driving the stirring shaft inside the hot washing tank. During equipment operation, the power transmitter and current transmitter are strictly aligned and synchronously collected in real time according to a preset high-frequency sampling interval to acquire instantaneous heating power and instantaneous stirring current. After acquiring the above instantaneous values, the difference between the instantaneous heating power at each high-frequency sampling moment and the instantaneous heating power at the adjacent previous high-frequency sampling moment is calculated, and this difference is directly used as the heating power variation at each high-frequency sampling moment; simultaneously, the difference between the instantaneous stirring current at each high-frequency sampling moment and the instantaneous stirring current at the adjacent previous high-frequency sampling moment is calculated, and this difference is directly used as the stirring current variation at each high-frequency sampling moment, thereby extracting the instantaneous load variation trend of the equipment.

[0020] It should be noted that, since historical measurement data from the adjacent previous sampling time cannot be obtained for difference calculation at the first high-frequency sampling time and the first low-frequency sampling time, this embodiment of the invention uses the actual measurement data collected at the first high-frequency sampling time and the first low-frequency sampling time as the initial reference state of the system, and formally calculates the above-mentioned various variables from the second high-frequency sampling time and the second low-frequency sampling time. This allows the data processing logic of the system to smoothly establish a basic reference at the initial startup stage, avoiding invalid calculations or errors caused by missing preceding data.

[0021] In one specific implementation of this invention, the preset low-frequency sampling period is exemplarily set to 10 minutes to ensure that the relevant analytical instruments have sufficient time to complete reagent mixing, reaction titration, and internal pipeline rinsing; the preset high-frequency sampling interval is exemplarily set to 1 second to ensure that the rising edge and peak shape of the mechanical resistance and heat load caused by the batch of plastic bottle flakes falling into the hot washing tank can be accurately captured. The implementer can adjust the preset low-frequency sampling period and preset high-frequency sampling interval according to the actual operating scale of the cleaning equipment, the instrument response performance, and the actual monitoring needs, which will not be further elaborated here.

[0022] Step S102: Based on the statistical distribution jumps of the heating power variation and stirring current variation over time, all continuous cleaning stages are divided; based on the consumption continuity relationship between each continuous cleaning stage and the previous continuous cleaning stage in terms of stirring current variation and free alkali concentration, the corresponding reaction residue level is determined.

[0023] Firstly, considering that different batches of materials falling into the cleaning tank will cause significant fluctuations in the instantaneous mechanical resistance and thermal load of the equipment, making the entire cleaning process exhibit step load characteristics closely related to the batch of materials, in order to accurately identify the independent reaction cycle triggered by the entry of each batch of materials and eliminate the monitoring blind spot where low-frequency sampling cannot align with the actual physical consumption rhythm, this embodiment of the invention first divides all continuous cleaning stages based on the statistical distribution jumps of heating power variation and stirring current variation in time sequence; by dividing the continuous equipment operation process into independent analysis units according to the actual material load impact events, and subsequently performing periodic control adjustments based on the continuous cleaning stages, the time anchor point of alkali replenishment decision can be closely aligned with the physical process of actual material consumption, enabling the automatic alkali replenishment control system to actively match the actual chemical reaction and consumption rhythm of each batch of materials.

[0024] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the continuous cleaning stage includes: Within a preset historical window before each high-frequency sampling time, the power response intensity is determined based on the mean of all heating power variations; the heating power fluctuation is determined based on the standard deviation of all heating power variations; the current drive intensity is determined based on the mean of all stirring current variations; and the stirring current fluctuation is determined based on the standard deviation of all stirring current variations.

[0025] In one specific implementation of this invention, the preset historical window time is set to 30 seconds for example. For high-frequency sampling moments when the initial state is less than 30 seconds away from the start, it is impossible to construct a complete preset historical window. In this case, all heating power variations and all stirring current variations within the first 30 seconds of system operation can be used for unified calculation to determine the power response intensity, heating power fluctuation, current drive intensity and stirring current fluctuation of each high-frequency sampling moment within the first 30 seconds, so as to ensure a smooth transition of the feature extraction process in the initial stage of startup.

[0026] Obtain the characteristic rise time; the corresponding values ​​of the characteristic rise time in power response intensity, heating power fluctuation, current drive intensity and stirring current fluctuation are all greater than the previous high-frequency sampling time; obtain the stage segment time; the previous high-frequency sampling time of the stage segment time is not the characteristic rise time, and the subsequent preset number of high-frequency sampling times are all characteristic rise times; divide the operation process into all continuous cleaning stages with the stage segment time as the stage interval.

[0027] In the aforementioned stage division process, the mean values ​​of heating power and stirring current variations reflect the baseline level of equipment load within a local time period, while their standard deviation characterizes the intensity of load fluctuations within that local time period. When a new batch of plastic bottles enters the washing tank, it inevitably disrupts the original steady state, leading to a significant increase in the baseline load and severe oscillations in operating conditions. Therefore, it is required that the baseline levels and fluctuations of power and current are synchronously greater than the previous moment to effectively capture the systematic jumps in the joint statistical distribution of electromechanical signals, thereby initially identifying the characteristic rise moment indicating material impact. Furthermore, considering the inevitable existence of power grid fluctuations or occasional disturbances in mechanical operation at industrial sites, these brief disturbances may also cause instantaneous increases in parameters. To eliminate stage misdivisions caused by such high-frequency noise, this embodiment does not directly use isolated parameter rises as stage boundaries, but instead seeks a precise jump starting point, i.e., before which the system is in a stable state, but from this point onwards, the subsequent preset number of high-frequency sampling moments all exhibit characteristic rises. This continuous and stable increase in multidimensional characteristics physically represents the large-scale entry of a new batch of materials. Therefore, this starting point is established as the stage segmentation moment, which serves as the precise physical boundary for dividing the continuous cleaning stage.

[0028] In one specific implementation of this invention, the preset quantity is exemplarily set to 6, which is used to characterize the stability of the continuous increase in equipment load characteristics caused by material impact. This value can be adjusted according to the specific implementation environment. If there are more high-frequency interference burrs in the field environment or the material falling process is more dispersed, the preset quantity should be set larger to improve the anti-interference ability.

[0029] It should be noted that since subsequent control logic relies on the chemical concentration at low-frequency sampling moments for boundary calculations and compensation feedback, this means that each continuous cleaning stage, as an independent analysis unit, must include at least one low-frequency sampling moment. Therefore, after the initial stage division, if a very short continuous cleaning stage contains no low-frequency sampling moments, this continuous cleaning stage without low-frequency sampling moments needs to be merged with the adjacent preceding continuous cleaning stage to ensure the integrity of the entire system's timing logic and concentration calculations.

[0030] After dividing all the continuous cleaning stages, this invention further considers that in the actual cleaning and degreasing process, the chemical reaction between the solution and the material surface requires a certain amount of time. When material enters in a concentrated manner at the end of the previous stage or the discharge is incomplete, the consumption reaction of some material is often not completed in time and is carried over to the current stage. This makes the calculation based solely on the current detection concentration underestimate the actual consumption burden. Therefore, this embodiment of the invention determines the degree of reaction residue that characterizes the incomplete reaction of the previous batch of material by determining the consumption gap of the solution based on the consumption continuity relationship between each continuous cleaning stage and the previous continuous cleaning stage in terms of the change in stirring current and free alkali concentration. The degree of reaction residue is used to quantify the delayed consumption that cannot be reflected in time by low-frequency sampling, providing a compensation basis for the subsequent determination of the alkali replenishment tendency coefficient.

[0031] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the degree of reaction residue includes: Each continuous cleaning stage is sequentially designated as the target cleaning stage, and the preceding continuous cleaning stage is designated as the reference cleaning stage. The segmentation time between the target cleaning stage and the reference cleaning stage is designated as the target segmentation time. The total concentration decrease of the target cleaning stage is determined based on the difference between the free alkali concentration at the last low-frequency sampling time of the reference cleaning stage and the free alkali concentration at the first low-frequency sampling time of the target cleaning stage. The total concentration decrease is allocated based on the variation in stirring current before and after the target segmentation time to determine the free alkali boundary concentration at the target segmentation time. Specifically: The final impact intensity is determined by summing the peak values ​​of all stirring current variations between the last low-frequency sampling moment of the reference cleaning stage and the target segmentation moment. The first impact intensity is determined by summing the peak values ​​of all stirring current variations between the target segmentation moment and the first low-frequency sampling moment of the target cleaning stage. The peak distribution weight at the target segmentation moment is determined based on the ratio between the final and first impact intensities. In this embodiment, the total transition impact intensity is determined based on the sum of the final and first impact intensities; the peak distribution weight at the target segmentation moment is determined based on the ratio between the final and transition impact intensities. It should be noted that if the total transition impact intensity is 0, it indicates that no material entered the hot washing tank between the two relevant low-frequency sampling moments. In this case, the system will directly set the peak distribution weight at the target segmentation moment to 0 to avoid a denominator of 0. Furthermore, regarding the specific method for obtaining the peak values ​​of stirring current variations, this embodiment uses the Automatic Multi-Scale Peak Detection (AMPD) algorithm to perform real-time scanning processing on the stirring current variation sequence obtained from high-frequency sampling to extract the corresponding peak values ​​of stirring current variations.

[0032] The total concentration decrease determined by the concentration difference between two low-frequency sampling times objectively reflects the total drug consumption within the sampling blind zone. However, drug consumption is not uniform but dynamically determined by the actual size of the plastic bottle flakes falling into the hot washing tank. Since the peak value of the stirring current variation is a direct physical representation of the sudden increase in mechanical resistance when the material enters, its accumulated value shows a high positive correlation with the drug consumption load within the corresponding time period. Therefore, this embodiment uses the target segment time as the physical time boundary, accumulates the peak values ​​of the stirring current variation within the sampling blind zone, and obtains the final impact intensity representing the end load of the previous stage and the first impact intensity representing the initial load of the new stage. By calculating the ratio of the two, the peak distribution weight obtained is essentially a precise quantification of the actual consumption weight of the total concentration decrease in the first half of the sampling blind zone.

[0033] Furthermore, based on the product of the peak distribution ratio and the total concentration decrease, the boundary concentration consumption allocation at the target segment time is determined. Based on the difference between the free alkali concentration at the last low-frequency sampling time of the reference cleaning stage and the boundary concentration consumption allocation, the free alkali boundary concentration at the target segment time is determined. Subtracting this boundary concentration consumption allocation from the free alkali concentration at the last low-frequency sampling time of the reference cleaning stage yields the free alkali boundary concentration corresponding to the physical segment boundary. This derivation method, compared to simple linear time interpolation, better reflects the actual dynamic reaction process of the material, enabling the accurate segmentation of continuous concentration consumption into independent physical reaction stages using the two derived concentration reference nodes.

[0034] Therefore, based on the difference between the free alkali boundary concentration at the target segment time and the free alkali boundary concentration at the corresponding next stage segment time, the actual consumption of the target cleaning stage is determined; the consumption conversion coefficient corresponding to the target cleaning stage is obtained; the expected consumption is determined based on the product of the consumption conversion coefficient and the cumulative value of all peak values ​​of stirring current fluctuations within the target cleaning stage; the reaction residue level of the target cleaning stage is determined based on the relative deviation between the expected consumption and the actual consumption; in this embodiment of the invention, the corresponding residual gap concentration is determined based on the difference between the expected consumption and the actual consumption; and the reaction residue level of the target cleaning stage is determined based on the ratio between the residual gap concentration and the expected consumption.

[0035] In one specific implementation of this invention, the process of obtaining the consumption conversion coefficient includes: obtaining all complete low-frequency sampling intervals within the target cleaning stage, wherein each complete low-frequency sampling interval is composed of two adjacent low-frequency sampling times, and both ends of the interval are located within the boundary of the target cleaning stage; calculating the difference between the free alkali concentrations at both ends of each complete low-frequency sampling interval to determine the corresponding interval concentration consumption; calculating the cumulative value of the peak values ​​of all stirring current fluctuations within each complete low-frequency sampling interval to determine the corresponding interval impact load; determining the corresponding interval conversion coefficient based on the ratio between the interval concentration consumption and the interval impact load of each complete low-frequency sampling interval; and determining the average of all interval conversion coefficients within the target cleaning stage as the consumption conversion coefficient corresponding to the target cleaning stage. It should be noted that if the interval concentration consumption is less than 0, it indicates an abnormal increase in free alkali concentration, which is highly likely due to external illegal dosing or sensor detection failure. In this case, the system will directly trigger an abnormal alarm and suspend the current automatic alkali replenishment process, prompting manual intervention for inspection.

[0036] After calculating the free alkali boundary concentrations at the start and end points of the two stages, the difference between the two concentrations represents the actual consumption of the chemical solution during the target cleaning stage. Since both ends of the complete low-frequency sampling interval are within the same target cleaning stage, its concentration change is not affected by material replacement across stages, and it can most purely reflect the chemical consumption rate under the current operating conditions. Therefore, its average value is used as a robust consumption conversion coefficient. Furthermore, multiplying this consumption conversion coefficient by the cumulative peak value of the stirring current variation reflecting the overall scale of the material during the entire target cleaning stage yields the theoretically expected consumption. When the theoretical expected consumption is greater than the actual consumption calculated by chemical analysis, the difference objectively reflects the amount of chemical reaction that could not be completed in time. By calculating the ratio, this can be converted into a relative proportion, definitively characterizing the degree of reaction residue that was not completed in time and remained in subsequent stages. It should be noted that if the expected consumption is 0, it means that no material enters during the target cleaning stage. In this case, the system will directly set the degree of reaction residue to 0 to avoid a denominator of 0 when calculating the ratio.

[0037] It should be noted that if the duration of the target cleaning stage is too short, resulting in fewer than two complete low-frequency sampling intervals being obtained within it, the median value of the consumption conversion coefficient of the most recently completed stage in the historical records can be retrieved as a substitute consumption conversion coefficient. If the system is in the initial startup stage and there is no historical stage data, the calculation and analysis of the target cleaning stage should be skipped. Furthermore, if an interval impact load is 0, the invalid interval should be directly removed. If a non-zero interval impact load cannot be found, it indicates that the cleaning equipment is in an abnormal state of no material feeding for a long time or is in a shutdown and idling state. In this case, the target alkali replenishment amount should be set to 0, and the automatic alkali replenishment operation should be suspended to avoid the danger of excessive concentration caused by blindly adding chemicals when there is no effective load.

[0038] Step S103: Based on the temporal evolution difference of the stirring current variation between each continuous cleaning stage and the previous continuous cleaning stage, determine the corresponding consumption trend intensity; based on the reaction residue level, consumption trend intensity, and cumulative concentration of ineffective byproducts, determine the corresponding alkali replenishment tendency coefficient.

[0039] Furthermore, considering the dynamic evolution of the degree of contamination and the continuous feeding rhythm between material batches, the consumption rate of the reagent in the current cleaning stage may accelerate or slow down compared to the previous stage. Therefore, this embodiment of the invention determines the corresponding consumption trend intensity based on the temporal evolution difference of the stirring current variation between each continuous cleaning stage and the previous continuous cleaning stage. Then, taking into account the feedforward compensation requirements for residual repayment and trend prediction within the low-frequency sampling period, as well as the possibility that excessive accumulation of ineffective by-products may prematurely trigger the operating limit boundary of equipment liquid replacement, the corresponding alkali replenishment tendency coefficient is determined based on the degree of reaction residue, the intensity of consumption trend, and the cumulative concentration of ineffective by-products. Through the above multi-dimensional comprehensive consideration, the automatic alkali replenishment control system can meet the reagent stabilization requirements while incorporating the cumulative risk of ineffective by-products into the constraint scope of alkali replenishment actions.

[0040] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the consumption trend strength includes: Based on the number of peak values ​​of all stirring current variations occurring within each continuous cleaning stage and the total stage duration, the corresponding stage impact frequency is determined; based on the relative deviation between the stage impact frequency of each continuous cleaning stage and the stage impact frequency of the previous continuous cleaning stage, the corresponding frequency trend component is determined; based on the average value of the peak values ​​of all stirring current variations within each continuous cleaning stage, the corresponding stage impact intensity is determined; based on the relative deviation between the stage impact intensity of each continuous cleaning stage and the stage impact intensity of the previous continuous cleaning stage, the corresponding intensity trend component is determined; based on the sum of the frequency trend component and the intensity trend component, the consumption trend intensity corresponding to each continuous cleaning stage is determined.

[0041] In this embodiment of the invention, a first difference is determined based on the difference between the stage impact frequency of each continuous cleaning stage and the stage impact frequency of the previous continuous cleaning stage; a frequency trend component is determined based on the ratio between the first difference and the stage impact frequency of the previous continuous cleaning stage; a second difference is determined based on the difference between the stage impact intensity of each continuous cleaning stage and the stage impact intensity of the previous continuous cleaning stage; and an intensity trend component is determined based on the ratio between the second difference and the stage impact intensity of the previous continuous cleaning stage.

[0042] In the quantification of the aforementioned trend intensity, the determined difference retains the physical direction of material load change: a positive difference indicates more frequent material entry or larger single-batch material entry, with accelerated chemical consumption; a negative difference indicates slower consumption. The further obtained relative deviation component comprehensively characterizes the evolution of the washing process from two orthogonal dimensions: material entry frequency and single-batch material load. This allows for the precise output of a consumption trend intensity with direction and amplitude through summation, providing a basis for subsequent prediction of whether alkali replenishment demand should increase or decrease. It should be noted that if the stage impact frequency or stage impact intensity of the previous continuous cleaning stage is 0, it indicates that the cleaning equipment was in a completely unloaded state with no material entry in the previous stage. In this case, it cannot be used as an effective benchmark to assess the relative rate of change of the trend. The system will directly identify the corresponding frequency trend component or intensity trend component as 0 to avoid a denominator of 0.

[0043] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the alkali supplementation tendency coefficient includes: The first low-frequency sampling time of the next continuous cleaning stage after each continuous cleaning stage is used as the corresponding control reference time. Based on the relative deviation between the preset upper limit of liquid exchange concentration and the cumulative concentration of ineffective byproducts at the control reference time, the corresponding liquid exchange constraint coefficient is determined. According to the sum of the degree of reaction residue and the intensity of consumption trend, the comprehensive compensation component of each continuous cleaning stage is determined. The product between the liquid exchange constraint coefficient and the comprehensive compensation component is positively correlated to determine the alkali replenishment tendency coefficient of each continuous cleaning stage.

[0044] This invention determines the corresponding capacity reserve concentration based on the difference between the preset upper limit of the liquid replacement concentration and the cumulative concentration of ineffective byproducts at the control reference time; and determines the corresponding liquid replacement constraint coefficient based on the ratio between the capacity reserve concentration and the preset upper limit of the liquid replacement concentration. The product of the liquid replacement constraint coefficient and the comprehensive compensation component, plus a real number 1, is used as the alkali replenishment tendency coefficient for each continuous cleaning stage. It should be noted that if the calculated alkali replenishment tendency coefficient is less than or equal to 0, it indicates that after considering the reaction residue and consumption trend, no additional reagent injection is needed under the current operating conditions. In this case, the system will directly set the final target alkali replenishment amount to 0 and skip subsequent adjustment steps.

[0045] This embodiment selects the first low-frequency sampling moment of the next continuous cleaning stage as the control reference moment. This ensures that after the previous physical cleaning stage is completely completed, the latest actual chemical detection status is used to execute the corresponding compensation decision, thereby ensuring that all dynamic compensation decisions are based on the most recent and reliable chemical concentration benchmark. The remaining capacity concentration directly reflects how much by-product capacity is left in the hot washing tank before the solution becomes unusable. By further calculating the ratio to determine the liquid replacement constraint coefficient, this remaining space is essentially transformed into a regulatory inhibition factor. Multiplying the comprehensive compensation component representing the feedforward regulation demand with the liquid replacement constraint coefficient reveals the core logic: the system's compensation for historical residues and its response to future trends must be strictly constrained by the current remaining liquid replacement space. When by-product accumulation is low and the remaining space is ample, the system can fully respond to dynamic compensation demands; however, when ineffective by-products approach the preset upper limit of the liquid replacement concentration, even if the material consumption gap is huge, the compensation impulse will be forcibly suppressed by the liquid replacement constraint coefficient, thereby preventing the alkali replenishment action from becoming a catalyst that accelerates the overall solution's obsolescence. Finally, the product is added to a real number 1 to establish a basic control baseline. When the system is in a steady-state condition with no historical residue and no load fluctuation, i.e., the comprehensive compensation component is 0, the product term returns to zero, the alkali supplementation tendency coefficient smoothly returns to 1, indicating that the system maintains the original basic dosage without making additional expansion or contraction adjustments.

[0046] In one specific implementation of this invention, the upper limit of the preset replacement solution concentration can be set according to the physicochemical properties of the alkaline cleaning agent used and the standard specifications of the degreasing process for bottle flakes. This setting is used to characterize the maximum critical concentration value corresponding to when the solution completely loses its cleaning activity or is sufficient to cause severe by-product precipitation.

[0047] Step S104: Determine the target alkali replenishment amount based on the alkali replenishment tendency coefficient and the free alkali concentration; automatically control the alkali replenishment of the plastic bottle flake washing equipment based on the target alkali replenishment amount.

[0048] After obtaining the alkali replenishment tendency coefficient, which characterizes the comprehensive control requirements of the current working conditions, the embodiment of the present invention finally determines the target alkali replenishment amount based on the alkali replenishment tendency coefficient and the free alkali concentration; and automatically controls the alkali replenishment of the plastic bottle flake cleaning equipment based on the target alkali replenishment amount. In this way, the automatic alkali replenishment command issued by the control system is no longer limited solely to the instantaneous detection concentration deviation, but integrates the feedforward control requirements based on the actual dynamic load of the material, effectively avoiding blind over-addition of chemicals and the accelerated accumulation of ineffective by-products, thereby maximizing the effective use cycle of the cleaning agent in the tank while maintaining the activity of the cleaning agent.

[0049] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the target alkali supplementation amount includes: The difference between the preset target concentration of free alkali and the free alkali concentration at the control reference time is calculated to determine the first concentration deviation. Based on the first concentration deviation, the preset effective volume of the cleaning equipment, and the preset free alkali concentration of the alkali replenishing agent, the corresponding basic alkali replenishment amount is determined. In this embodiment of the invention, the first concentration deviation is multiplied by the preset effective volume of the cleaning equipment, and the product is divided by the preset free alkali concentration of the alkali replenishing agent to determine the basic alkali replenishment amount. Based on the product between the basic alkali replenishment amount and the alkali replenishment tendency coefficient, the required alkali replenishment amount for each continuous cleaning stage is determined. The above-mentioned basic alkali replenishment amount only represents the theoretical volume required to make up for the current static concentration gap. By multiplying and correcting it by the alkali replenishment tendency coefficient, dynamic operating condition characteristics such as reaction residue and consumption trend derived from previous calculations can be injected into it, so that the corrected required alkali replenishment amount can truly match the actual dynamic consumption rhythm of the equipment. It should be noted that if the calculated first concentration deviation is less than or equal to 0, it means that the free alkali concentration in the tank has reached or exceeded the target requirement, and there is no need to replenish alkali. The system will directly set the target alkali replenishment amount to 0 and skip the calculation of the subsequent basic alkali replenishment amount and the maximum alkali replenishment volume.

[0050] Based on the preset upper limit of the liquid replacement concentration and the cumulative concentration of ineffective by-products at the control reference time, the limit alkali replenishment volume for each continuous cleaning stage is determined. Specifically: the difference between the preset upper limit of the liquid replacement concentration and the cumulative concentration of ineffective by-products at the control reference time is calculated to determine the second concentration deviation; the second concentration deviation is multiplied by the preset effective volume of the cleaning equipment, and the product is divided by the preset free alkali concentration of the alkali replenishment agent to determine the corresponding limit alkali replenishment volume. This limit alkali replenishment volume objectively quantifies the maximum absolute limit of the current chemical addition volume that the hot washing tank can accommodate without triggering forced liquid replacement. It should be noted that if the calculated second concentration deviation is less than or equal to 0, it indicates that the tank solution may be severely aged, causing the cumulative concentration of by-products to exceed the liquid replacement limit. In this case, an abnormal alarm is triggered, the current automatic alkali replenishment process is suspended, and manual intervention is requested for inspection.

[0051] Based on the minimum value between the required alkali replenishment amount and the maximum alkali replenishment volume, the target alkali replenishment amount for each continuous cleaning stage is determined. By taking the minimum value, a rigid physical safety boundary is essentially built at the execution end of the control system, ensuring that no matter how large the calculated required alkali replenishment amount is, the actual output final dosing command will never exceed the by-product capacity limit.

[0052] In one specific implementation of this invention, the free alkali concentration of the preset alkali replenishing agent is obtained in advance by chemically measuring the fresh alkali replenishing solution to be added. The effective volume of the preset cleaning equipment is obtained in advance based on the physical structure dimensions of the internal hot washing tank of the plastic bottle flake cleaning equipment and the rated working liquid level, which will not be further elaborated here.

[0053] Upon reaching the control reference time, i.e., the first low-frequency sampling moment of the newly initiated continuous cleaning phase, the control system will directly generate a corresponding dosing drive command based on the target alkali replenishment amount calculated at that moment, and send the command to the automatic alkali replenishment actuator of the plastic bottle flake cleaning equipment. Upon receiving the command, the automatic alkali replenishment actuator will precisely inject a volume of alkali replenishing solution corresponding to the target alkali replenishment amount into the hot washing tank of the cleaning equipment in real time, thus completing the automatic alkali replenishment operation for this control cycle. With the continuous operation of the cleaning equipment and the continuous updating of various high- and low-frequency monitoring data, the control system will repeatedly execute the above-mentioned continuous cleaning phase division and data analysis process, and trigger and execute the corresponding alkali replenishment action in real time at the first low-frequency sampling moment of each subsequent new continuous cleaning phase, thereby achieving adaptive dynamic maintenance and closed-loop control of the chemical activity of the cleaning solution.

[0054] In summary, an online monitoring and feedback control method for plastic bottle flake washing equipment divides continuous washing stages based on the statistical distribution jumps of heating power and stirring current variations, effectively capturing the dynamic changes caused by material entry. By analyzing the consumption continuity relationship and temporal evolution differences between adjacent stages, the degree of reaction residue and the intensity of consumption trend are determined respectively, accurately quantifying the actual consumption changes and residual effects within the low-frequency sampling interval. Furthermore, the method comprehensively determines the alkali replenishment tendency coefficient by combining the cumulative concentration of ineffective byproducts, and determines the target alkali replenishment amount based on this coefficient and the free alkali concentration. This method incorporates the real dynamic consumption rhythm and byproduct accumulation constraints into the control process, effectively avoiding excessive alkali replenishment and blind accumulation of byproducts caused by relying solely on low-frequency concentration deviations. It solves the technical problem of low accuracy in existing technologies that rely on low-frequency sampling concentration deviations for automatic alkali replenishment control, making the automatic alkali replenishment control of this application more accurate.

Claims

1. An online monitoring and feedback control method for a plastic bottle flake washing equipment, characterized in that, The method includes: During the operation of the plastic bottle flake washing equipment, the free alkali concentration and cumulative concentration of ineffective byproducts of the washing equipment at each low-frequency sampling time, as well as the heating power variation and stirring current variation at each high-frequency sampling time are obtained. Based on the statistical distribution jumps of heating power variation and stirring current variation over time, all continuous cleaning stages are divided; based on the consumption continuity relationship between each continuous cleaning stage and the previous continuous cleaning stage in terms of stirring current variation and free alkali concentration, the corresponding reaction residue degree is determined. Based on the temporal evolution difference of the stirring current variation between each continuous cleaning stage and the previous continuous cleaning stage, the corresponding consumption trend intensity is determined; based on the reaction residue level, the consumption trend intensity, and the cumulative concentration of ineffective byproducts, the corresponding alkali replenishment tendency coefficient is determined. Based on the alkali replenishment tendency coefficient and the free alkali concentration, the target alkali replenishment amount is determined; based on the target alkali replenishment amount, the plastic bottle flake washing equipment is automatically controlled to replenish alkali.

2. The online monitoring and feedback control method for a plastic bottle flake washing equipment according to claim 1, characterized in that, The process of obtaining the continuous cleaning phase includes: Within a preset historical window before each high-frequency sampling moment, the power response intensity is determined based on the mean of all heating power variations; the heating power fluctuation is determined based on the standard deviation of all heating power variations; the current drive intensity is determined based on the mean of all stirring current variations; and the stirring current fluctuation is determined based on the standard deviation of all stirring current variations. The characteristic rise time is obtained; the values ​​corresponding to the characteristic rise time in terms of power response intensity, heating power fluctuation, current drive intensity and stirring current fluctuation are all greater than the previous high-frequency sampling time. Obtain the stage segmentation time; the preceding high-frequency sampling time of the stage segmentation time is not the feature rise time, and the subsequent consecutive preset number of high-frequency sampling times are all feature rise times; The process is divided into all continuous cleaning stages, with the stage segmentation time as the stage interval.

3. The online monitoring and feedback control method for a plastic bottle flake washing equipment according to claim 2, characterized in that, The process for obtaining the degree of reaction residue includes: Each consecutive cleaning stage is sequentially designated as the target cleaning stage, and the preceding consecutive cleaning stage is designated as the reference cleaning stage; the time of segmentation between the target cleaning stage and the reference cleaning stage is designated as the target segmentation time. The total concentration decrease in the target cleaning stage is determined based on the difference between the free alkali concentration at the last low-frequency sampling moment of the reference cleaning stage and the free alkali concentration at the first low-frequency sampling moment of the target cleaning stage. The total concentration decrease is allocated based on the change in stirring current before and after the target segment time, and the free alkali boundary concentration at the target segment time is determined. The actual consumption of the target cleaning stage is determined based on the difference between the free alkali boundary concentration at the target segment time and the free alkali boundary concentration at the corresponding next stage segment time. Obtain the consumption conversion coefficient corresponding to the target cleaning stage; determine the expected consumption based on the product of the consumption conversion coefficient and the cumulative value of the peak values ​​of all stirring current variations within the target cleaning stage; determine the degree of reaction residue in the target cleaning stage based on the relative deviation between the expected consumption and the actual consumption.

4. The online monitoring and feedback control method for a plastic bottle flake washing equipment according to claim 3, characterized in that, The process of obtaining the free base boundary concentration includes: The final impact intensity is determined by summing the peak values ​​of all stirring current variations between the last low-frequency sampling moment of the reference cleaning phase and the target segmentation moment; the first impact intensity is determined by summing the peak values ​​of all stirring current variations between the target segmentation moment and the first low-frequency sampling moment of the target cleaning phase; and the peak distribution weight at the target segmentation moment is determined based on the ratio between the final impact intensity and the first impact intensity. The boundary concentration consumption allocation at the target segment time is determined based on the product of the peak distribution ratio and the total concentration decrease; the free alkali boundary concentration at the target segment time is determined based on the difference between the free alkali concentration at the last low-frequency sampling time of the reference cleaning stage and the boundary concentration consumption allocation.

5. The online monitoring and feedback control method for a plastic bottle flake washing equipment according to claim 1, characterized in that, The process of obtaining the intensity of the consumption trend includes: Based on the number of peak values ​​of all stirring current variations that occur in each continuous cleaning stage and the total duration of the stage, the corresponding stage impact frequency is determined; based on the relative deviation between the stage impact frequency of each continuous cleaning stage and the stage impact frequency of the previous continuous cleaning stage, the corresponding frequency trend component is determined. The corresponding stage impact intensity is determined based on the average of the peak values ​​of all stirring current variations within each continuous cleaning stage; the corresponding intensity trend component is determined based on the relative deviation between the stage impact intensity of each continuous cleaning stage and the stage impact intensity of the previous continuous cleaning stage. The consumption trend intensity corresponding to each continuous cleaning stage is determined based on the sum of the frequency trend component and the intensity trend component.

6. The online monitoring and feedback control method for a plastic bottle flake washing equipment according to claim 1, characterized in that, The process of obtaining the alkali supplementation tendency coefficient includes: The first low-frequency sampling time of the next continuous cleaning stage of each continuous cleaning stage is used as the corresponding control reference time. Based on the relative deviation between the preset upper limit of the liquid replacement concentration and the cumulative concentration of ineffective byproducts at the control reference time, the corresponding liquid replacement constraint coefficient is determined. Based on the sum of the residual reaction level and the intensity of the consumption trend, a comprehensive compensation component for each continuous cleaning stage is determined; the product of the liquid exchange constraint coefficient and the comprehensive compensation component is positively correlated to determine the alkali replenishment tendency coefficient for each continuous cleaning stage.

7. The online monitoring and feedback control method for a plastic bottle flake washing equipment according to claim 6, characterized in that, The process for obtaining the target alkali supplementation amount includes: Calculate the difference between the preset target concentration of free alkali and the concentration of free alkali at the control reference time, and determine the first concentration deviation; Based on the first concentration deviation, the preset effective volume of the cleaning equipment, and the preset free alkali concentration of the alkali replenishing agent, the corresponding basic alkali replenishment amount is determined; wherein, the basic alkali replenishment amount is positively correlated with the first concentration deviation and the preset effective volume of the cleaning equipment, and negatively correlated with the preset free alkali concentration of the alkali replenishing agent; Based on the product of the basic alkali replenishment amount and the alkali replenishment tendency coefficient, the required alkali replenishment amount for each continuous cleaning stage is determined. Based on the preset upper limit of the liquid replacement concentration and the cumulative concentration of ineffective byproducts at the control reference time, the limit alkali replenishment volume for each continuous cleaning stage is determined. The target alkali replenishment amount for each continuous cleaning stage is determined based on the minimum value between the required alkali replenishment amount and the maximum alkali replenishment volume.

8. The online monitoring and feedback control method for a plastic bottle flake washing equipment according to claim 7, characterized in that, The process of obtaining the maximum alkali replenishment volume includes: Calculate the difference between the preset upper limit of the replacement solution concentration and the cumulative concentration of ineffective byproducts at the control reference time to determine the second concentration deviation; based on the second concentration deviation, the preset effective volume of the cleaning equipment, and the preset free alkali concentration of the alkali replenishing agent, determine the corresponding limit alkali replenishment volume; wherein, the limit alkali replenishment volume is positively correlated with the second concentration deviation and the preset effective volume of the cleaning equipment, and negatively correlated with the preset free alkali concentration of the alkali replenishing agent.

9. The online monitoring and feedback control method for a plastic bottle flake washing equipment according to claim 4, characterized in that, The process of obtaining the peak distribution weight includes: The total transition impact intensity is determined based on the sum of the final impact intensity and the initial impact intensity. The peak distribution ratio at the target segment time is determined based on the ratio between the final impact intensity and the sum of the transition impact intensities.

10. The online monitoring and feedback control method for a plastic bottle flake washing equipment according to claim 6, characterized in that, The process of obtaining the fluid exchange constraint coefficient includes: The corresponding capacity margin concentration is determined based on the difference between the preset upper limit of the fluid replacement concentration and the cumulative concentration of ineffective byproducts at the control reference time. The corresponding fluid replacement constraint coefficient is determined based on the ratio between the remaining capacity concentration and the preset upper limit of the fluid replacement concentration.

Images