Polyacrylate emulsion post adjustment stage thickener addition with in-line viscosity control system

Abstract

The present application relates to the technical field of chemical production regulation, and discloses a thickener adding and online viscosity regulation system in the post-adjustment stage of polyacrylic acid emulsion. The system comprises an overview construction module, a coefficient generation module, a rate optimization module and an execution control module. The overview construction module creates a production overview based on user settings, including the connection relationship of the reaction tank body and the parameters of the thickener injection node; the coefficient generation module collects real-time stirring state data and tank body reaction data, respectively constructs a stirring state model and a tank body reaction model, and outputs stirring state coefficients and tank body reaction coefficients; the rate optimization module obtains the difference between the current viscosity measurement value and the target viscosity, combines the tank body reaction coefficients, the stirring state coefficients and the viscosity difference to construct a rate optimization model, and outputs the target thickener injection rate; and the execution control module drives the thickener injection device according to the target thickener injection rate. The system improves the accuracy and stability of viscosity regulation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of chemical production regulation, in particular to a thickener adding and online viscosity regulation system for the post-adjustment stage of polyacrylic emulsion. BACKGROUND

[0002] In the production process of polyacrylic emulsion, the viscosity regulation in the post-adjustment stage plays a crucial role in the final performance of the product, and the amount and rate of thickener addition are the core factors affecting the viscosity regulation effect. At present, most production enterprises in the industry still rely on traditional experience-based operation mode in this stage, and the operating personnel manually adjust the amount of thickener injection according to the offline detected viscosity data. This method has obvious hysteresis. Offline detection needs multiple links such as sampling, testing and analysis, and the whole process takes a long time, so the detection result cannot reflect the current state of the reaction tank in real time. The adjustment of thickener addition often lags behind the actual reaction demand, which may cause large viscosity fluctuations.

[0003] The traditional regulation method does not consider the correlation between the stirring state and the tank reaction environment. The viscosity change of polyacrylic emulsion is not only related to the amount of thickener added, but also closely related to the stirring rate, stirring uniformity, material temperature, pH value and other factors of the reaction tank. In the existing operation, the stirring state data and the tank reaction data are recorded separately, and there is no effective linkage analysis mechanism, so it is difficult to accurately capture the internal law of viscosity change under different working conditions. When the stirring state is unstable or the tank reaction parameters fluctuate, the thickener addition rate cannot be adjusted in time, which may cause local thickener aggregation, uneven reaction and other problems, affecting the stability and consistency of the emulsion.

[0004] The setting of thickener injection nodes lacks systematic planning. The structure parameters, material capacity and pipeline connection relationship of different reaction tanks are different. If the position, number and injection pressure of the thickener injection point are not reasonably set, the dispersion effect of the thickener in the emulsion will be poor, which will further increase the difficulty of viscosity regulation. Some enterprises have tried to introduce simple automatic injection devices, but due to the lack of dynamic analysis of real-time data and model support, the injection rate adjustment still relies on fixed parameters, which cannot adapt to the dynamic changes in the production process, resulting in large quality differences between product batches, unstable qualified rate, and increased production rework rate and raw material loss. SUMMARY

[0005] The purpose of the present application is to provide a thickener adding and online viscosity regulation system for the post-adjustment stage of polyacrylic emulsion to solve the problems raised in the background.

[0006] To achieve the above purpose, the present application provides a thickener adding and online viscosity regulation system for the post-adjustment stage of polyacrylic emulsion, which comprises:

[0007] The overview construction module is configured to create a production overview based on user settings, the production overview including a connection relationship of a reaction tank body and a thickening agent injection node parameter;

[0008] The coefficient generation module is configured to collect real-time stirring state data and tank reaction data, construct a stirring state model based on the stirring state data to output a stirring state coefficient, and construct a tank reaction model based on the tank reaction data to output a tank reaction coefficient;

[0009] The rate optimization module is configured to obtain a viscosity difference value between a current viscosity measurement value and a target viscosity, construct a rate optimization model based on the tank reaction coefficient, the stirring state coefficient and the viscosity difference value, and output a target thickening agent injection rate;

[0010] The execution control module is configured to drive a thickening agent injection device according to the target thickening agent injection rate.

[0011] Preferably, the coefficient generation module constructs a stirring state model based on stirring state data to output a stirring state coefficient, including: calling a preset stirring speed and blade torque mapping relationship, matching current stirring speed data to obtain a theoretical torque reference value;

[0012] calculating a torque deviation amount between an actual torque measurement value and the theoretical torque reference value;

[0013] determining a stirring state coefficient according to the torque deviation amount and a preset torque deviation level table, wherein the stirring state coefficient decreases correspondingly when the torque deviation amount increases.

[0014] Preferably, the coefficient generation module constructs a tank reaction model based on tank reaction data to output a tank reaction coefficient, including: obtaining reaction temperature time series data and pH value time series data;

[0015] extracting fluctuation characteristics from the reaction temperature time series data to obtain a temperature fluctuation index, and extracting drift characteristics from the pH value time series data to obtain a pH drift index;

[0016] inputting the temperature fluctuation index and the pH drift index into a preset tank reaction coefficient conversion table to match and output the tank reaction coefficient.

[0017] Preferably, the rate optimization module constructs a rate optimization model based on the tank reaction coefficient, the stirring state coefficient and the viscosity difference value, including: calling a preset coefficient influence weight reference table, determining a first adjustment weight according to the tank reaction coefficient, and determining a second adjustment weight according to the stirring state coefficient;

[0018] calculating an initial thickening agent injection rate based on the viscosity difference value and a preset basic rate conversion relationship;

[0019] The initial thickening agent injection rate is combined with the first adjustment weight and the second adjustment weight to generate the target thickening agent injection rate.

[0020] Preferably, the system further comprises a feedback calibration module; the feedback calibration module is configured to collect an actual viscosity variation gradient after the target thickening agent injection rate is executed;

[0021] When the actual viscosity variation gradient deviates from a preset expected gradient range, a model calibration instruction is generated;

[0022] The coefficient generation module updates a torque deviation level table in the stirring state model or updates a tank reaction coefficient conversion table in the tank reaction model in response to the model calibration instruction.

[0023] Preferably, the feedback calibration module is further configured to count a viscosity control overshoot number within a preset period of time;

[0024] When the viscosity control overshoot number reaches a preset threshold, a weight adjustment instruction is generated;

[0025] The rate optimization module resets the first adjustment weight or the second adjustment weight in the coefficient influence weight table in response to the weight adjustment instruction.

[0026] Preferably, the execution control module drives a thickening agent injection device according to the target thickening agent injection rate, including: obtaining current liquid level height and pipeline pressure data of a thickening agent storage tank; calculating an injection rate compensation amount according to the current liquid level height and a preset liquid level compensation relationship;

[0027] calculating an injection rate decay amount according to the pipeline pressure data and a preset pressure decay relationship;

[0028] The target thickening agent injection rate is superimposed with the injection rate compensation amount and subtracted by the injection rate decay amount to generate a final execution rate output to the thickening agent injection device.

[0029] Preferably, the system further comprises a priority processing module; the priority processing module is configured to monitor an emergency state signal of a reaction tank;

[0030] When a temperature overrun signal or a pressure overrun signal is received, an injection interruption instruction is generated to override the final execution rate;

[0031] When the emergency state signal is released, the output of the final execution rate is restored.

[0032] Preferably, when the feedback calibration module collects the actual viscosity variation gradient, it includes: recording continuous sampling values of a viscosity sensor within a unit time;

[0033] calculating an absolute variation sequence of the adjacent sampling values;

[0034] performing a moving average process on the absolute variation sequence to obtain the actual viscosity variation gradient.

[0035] Preferably, the production overview created by the overview construction module contains the connection relationship of the reaction tank body, including: identifying the fluid communication path of the main reaction tank and the buffer tank;

[0036] defining the topological position of the thickener injection point in the fluid communication path;

[0037] associating the viscosity monitoring device identifier and the thickener valve control identifier of each node.

[0038] Compared with the prior art, the present application has the following advantages:

[0039] The thickener addition in the post-adjustment stage of the polyacrylic acid emulsion and the online viscosity control system realize intelligent and fine management of thickener addition and viscosity control through multi-module cooperation. The overview construction module creates a production overview containing the connection relationship of the reaction tank body and the parameters of the thickener injection node based on user settings, enabling the operator to intuitively grasp the key node information of the entire production process and clearly understand the association between each tank, the specific position of the thickener injection, and the parameter setting, providing a clear global perspective for subsequent control and avoiding the blindness of control caused by scattered information.

[0040] The coefficient generation module breaks the data isolation in traditional control by collecting real-time stirring state data and tank reaction data, respectively constructing stirring state models and tank reaction models, and outputting corresponding coefficients. The stirring state data covers key indicators such as stirring rate and stirring power, and the tank reaction data includes parameters such as material temperature, concentration, and pH value. The construction process of the model fully explores the potential correlation between these data and viscosity changes, and the output coefficients can truly reflect the influence of the current stirring effect and tank reaction environment on viscosity, providing a basis for the optimization of thickener injection rate that fits the actual production state.

[0041] The rate optimization module takes the difference between the current viscosity measurement value and the target viscosity as the core reference, combines the tank reaction coefficient and the stirring state coefficient to construct a rate optimization model, and realizes dynamic calculation of the thickener injection rate. This model does not rely on fixed formulas or empirical values, but continuously adjusts the calculation logic according to real-time data changes, so that the output target thickener injection rate can accurately adapt to the current reaction state. When the stirring state changes cause the material mixing efficiency to change, or the tank reaction parameters fluctuate and affect the thickener effect, the rate optimization module can respond in time, avoiding the problems of viscosity overshoot or insufficient adjustment caused by traditional fixed rate injection.

[0042] The execution control module drives the injection device according to the target thickening agent injection rate, ensuring accurate execution of the control instruction. The entire process from data acquisition, model calculation to instruction execution forms a closed loop, reducing errors caused by human intervention, making the thickening agent addition process more stable and reliable. Whether it is small-scale trial production or large-scale continuous production, the system can maintain consistent control accuracy, adapt to viscosity control requirements under different production scales, and improve the stability of the polyacrylic acid emulsion post-adjustment stage production process and the consistency of product quality. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 A timing chart of the polyacrylic acid emulsion post-adjustment stage thickening agent addition and online viscosity control system according to the present application;

[0044] Figure 2 A flowchart for constructing the stirring state model of the coefficient generation module;

[0045] Figure 3 A flowchart for constructing the tank reaction model of the coefficient generation module;

[0046] Figure 4 A flowchart for executing the thickening agent injection rate of the execution control module. DETAILED DESCRIPTION

[0047] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of the present application.

[0048] Please refer to Figure 1 The present application provides a polyacrylic acid emulsion post-adjustment stage thickening agent addition and online viscosity control system, which comprises:

[0049] The overview construction module is responsible for receiving and processing user configuration instructions, which are set by the user through the operation interface to involve the reaction tank body of the production process, including the main reaction tank, the intermediate buffer tank, and the like, as well as the fluid transport path, the pipeline connection mode and the flow direction among them. At the same time, the user needs to set key parameters, especially the thickener injection point information located on the selected pipeline or tank body, such as the preset initial injection rate range, the associated viscosity monitoring point identifier, and the like. The module integrates these information to form a structured, machine-readable production overview, which clearly describes the physical layout and initial setting parameters of each functional unit in the entire post-adjustment stage. The coefficient generation module runs continuously to collect real-time operation data from the stirring device and the inside of the tank body. Specifically, it includes real-time stirring state data of the stirring paddle, mainly including real-time speed and motor torque measurement values, and tank reaction data in the reaction tank, mainly including temperature and pH value sequences that change continuously over time. The core function of this module is to construct special calculation models according to the two types of data. For stirring data, the module will calculate the stirring state coefficient reflecting the current stirring homogeneity efficiency according to the preset processing rules. For the reaction tank data, the module will analyze the change characteristics of temperature and pH, and calculate the tank reaction coefficient reflecting the stability of the mixing and reaction in the tank. Both of these coefficients are dynamic numerical values.

[0050] The rate optimization module compares the current viscosity measurement value of the reaction system obtained by the online viscosity sensor with the target viscosity set by the user to calculate the viscosity difference, which is usually represented as the difference between the target value and the measured value. The module obtains the tank reaction coefficient and the stirring state coefficient output by the coefficient generation module, combines the aforementioned viscosity difference, and calculates and decides through the preset optimization algorithm model. The output result of the model is the dynamically adjusted target thickener injection rate, which aims to effectively narrow the viscosity difference according to the current process state. The execution control module receives the target thickener injection rate instruction issued by the rate optimization module, drives the physical thickener injection device such as a quantitative pump and a regulating valve through a control signal digital communication instruction, accurately delivers the thickener to the specified injection point at the calculated rate, and further controls the viscosity of the reaction system to approach the target value.

[0051] Example 1: see Figure 2, the detailed calculation process of the stirring state coefficient, specifically including real-time data acquisition, benchmark matching, deviation calculation, and complete technical solutions for coefficient mapping. The system continuously acquires the operating parameters of the stirring equipment through a high-speed data acquisition interface. The core data includes the real-time rotational speed value of the stirring paddle fed back by the encoder of the driving motor, usually measured in revolutions per minute or radians per second, and the actual torque measurement value output by the torque sensor installed on the driving shaft, usually measured in Newton per meter. These raw data are transmitted to the central processing unit through an analog-to-digital conversion module or a direct digital interface and stored in a circular buffer to ensure data timeliness and processing continuity. The construction of the stirring state model relies on a preset, non-volatile storage device such as an EEPROM or a database that stores a "stirring rotational speed and paddle torque mapping relationship" dataset. This dataset is established based on a large number of equipment no-load calibration tests and load characteristic tests conducted in uniform standard fluids such as pure water and glycerol standard solution. The required steady-state torque range or average value of the driving shaft at a specific rotational speed setpoint is recorded, thereby forming a correspondence between the rotational speed value and the theoretical torque benchmark value. This mapping relationship can be presented as a discrete two-dimensional lookup table form, with each row storing a rotational speed interval and the corresponding theoretical torque benchmark value, or as a continuous function form stored in the form of a polynomial function containing the first-order and second-order coefficients of rotational speed. The system immediately starts a matching query after receiving new real-time stirring rotational speed data each time. If the mapping relationship exists in the form of a lookup table, the current measured rotational speed value is used to determine the preset rotational speed interval it falls into, such as the rotational speed table header divided into: 0-100 , 101-200 , …, and the corresponding theoretical torque benchmark value 8.2 Newton per meter in the 150 rpm interval is extracted. If the mapping relationship is in the form of a function, the current rotational speed value is directly substituted into the preset formula containing the first-order and second-order coefficients of rotational speed to calculate the real-time theoretical torque benchmark value. This theoretical torque benchmark value represents the expected torque load level that should be reached under the condition of uniform and stable fluid properties and normal operation of the stirring system at a given rotational speed, serving as an objective basis for subsequent comparison.

[0052] After obtaining the theoretical torque reference value, the system computing unit performs the key deviation calculation. This unit reads the actual torque measurement value at the same time or strictly synchronized with the time stamp of the theoretical torque reference value calculation from the data buffer to ensure the consistency and time synchronization of the comparison reference. The actual torque measurement value has usually been processed by noise filtering such as low-pass filtering or moving average processing. The calculation module performs an algebraic subtraction operation: the torque deviation amount is equal to the actual torque measurement value minus the theoretical torque reference value. The deviation amount calculation result is a positive or negative numerical value unit Newton per meter. A positive deviation amount indicates that the actual torque load on the drive shaft is higher than the expected value under the standard uniform fluid state, implying that the stirring paddle encounters greater stirring resistance, which may be caused by abnormal viscosity increase, suspended particle sedimentation and accumulation, local gelation, or tank wall material hanging, etc. A negative deviation amount indicates that the actual torque is lower than expected, which may be caused by too low material viscosity, insufficient filling of the fluid to the paddle area, transmission slip, or sensor drift, etc. The absolute value of the deviation amount directly quantifies the degree of deviation of the actual stirring state from the ideal mixing working condition. The calculation result is input into the pre-set "torque deviation level table" to complete the coefficient mapping process. This level table is the core static configuration data, also stored in the non-volatile memory, and its essence is to discretize the continuous torque deviation amount absolute value range into a finite number of intervals, and specify a unique stirring state coefficient value for each interval. The table structure usually contains two or more columns, for example, the first column defines the lower and upper limits of the torque deviation amount absolute value unit Newton per meter, and the second column stores the corresponding coefficient value. The core logic rule of the table design is: the larger the absolute value of the torque deviation amount, the smaller the stirring state coefficient value mapped. For example, a specific level table entry can be designed as follows: deviation amount interval 0.0 to 1.0 Newton per meter corresponds to coefficient 1.0; 1.0 to 2.5 Newton per meter corresponds to coefficient 0.8; 2.5 to 4.0 Newton per meter corresponds to coefficient 0.6; and 4.0 or more corresponds to coefficient 0.4. This reflects the objective physical law that the stirring efficiency decreases with the increase of load resistance: the larger the deviation, the stronger the interference of the stirring effect by adverse factors, and the model represents this negative effect by assigning it a smaller coefficient value. After calculating the torque deviation amount, the system takes its absolute value and uses it as an input index to match the level table. The matching mechanism is: check one by one which pre-defined deviation interval the absolute value belongs to, the lower limit is less than or equal to the absolute value and the absolute value is less than the upper limit, once successfully matched to a certain interval row, read and output the associated stirring state coefficient value from the column corresponding to the row. Boundary condition processing: when the absolute value is exactly equal to the lower limit of a certain interval, the system clearly belongs to the corresponding interval according to the pre-defined rule. If the absolute value exceeds the maximum upper limit value listed in the table, it is defaulted to match the last interval defined or take the minimum coefficient value corresponding to that interval.

[0053] The calculation and output process of the stirring state coefficient is automatic and continuous throughout the whole process of thickener addition control. The entire process highly depends on the real-time and accurate speed and torque signals provided by the hardware sensors. The signal acquisition path has a certain physical connection relationship and signal processing protocol: physical sensor, photoelectric encoder, strain gauge or magnetic elastic torque meter → signal conditioning circuit, amplification, filtering, isolation → AD sampling module or digital communication interface such as RS485 / CAN → data acquisition driver program → shared memory / message queue → central processing unit calculation logic module. The core mapping basis "stirring speed and paddle torque mapping relationship" is not permanent and unchangeable. The system design allows data updating and calibration through the maintenance interface engineering parameter configuration interface, for example, according to the mechanical wear characteristics change after long-time running of the equipment, or re-calibration after replacing the stirring paddle type, and correcting the original reference value. This updating operation is independent of the real-time control cycle, and is usually carried out in the non-running state or service mode of the system. The physical storage structure of the lookup table can be a simple key-value pair table in the embedded database, or a structured array in the configuration file. In order to prevent runtime errors, the system has built-in checks for the validity of input parameters, such as the speed must be within the allowed range of the device, and the torque value must be within the range of the sensor. During operation, the system log subsystem automatically records the complete data packet of each coefficient calculation, including timestamp, real-time speed, theoretical torque reference value source description, actual torque measurement value, calculated torque deviation, matched deviation interval description, and finally output stirring state coefficient value. These log entries have precise time markers and association identifiers, forming traceable analysis records, which facilitate problem diagnosis and model optimization. On the operator monitoring interface, in addition to displaying the coefficient value in real time, the corresponding qualitative state indication can also be displayed, such as when the coefficient is greater than or equal to 0.9, display green "good", 0.7 to 0.9 display yellow "medium", and less than 0.7 display red "poor", providing a visual reference for the operator's situational awareness. This implementation converts the mechanical load characteristics of the stirring system into a standardized coefficient that represents the current mixing physical environment, especially the influence of the rheological state on the uniform distribution efficiency of the material, through objective and quantitative steps. This coefficient is called by the downstream rate optimization module as one of the important inputs for adjusting the thickener injection rate weight, and its value fluctuation directly affects the dynamic adjustment process of the final control target. This method is a specific implementation way of converting mechanical load characteristics into process observable control signals.

[0054] Example 2: see Figure 3The present embodiment covers the generation of the tank reaction coefficient and the optimization of the target thickener injection rate. The system continuously acquires raw signals from the temperature sensor array and the pH sensor installed in the reaction tank. The temperature sensor outputs the medium temperature measurements at fixed sampling intervals, forming a time series dataset, which constitutes the reaction temperature time series data. The pH sensor synchronously outputs the measurement sequence, which constitutes the pH value time series data. The processing of the temperature time series data focuses on fluctuation feature extraction: the system sets a sliding time window, for example, containing the latest 60 sampling points, corresponding to 1 minute of data. Within the window, the statistical dispersion index of the temperature data is calculated: the temperature fluctuation index. The calculation of this index uses the standard deviation method, specifically, the square root of the average of the sum of the squares of the deviations of each temperature measurement from its average value within the window. This value quantifies the fluctuation amplitude of the temperature in a short period of time, and the larger the value, the worse the temperature stability. The processing of the pH time series data focuses on drift feature extraction: within the same sliding time window, linear regression analysis is performed on the pH value sequence. The linear fitting slope of the pH value change with time within the window is calculated, and the absolute value of the slope is taken as the pH drift index. The larger the value, the more significant the continuous upward or downward trend of the pH value within the window period, and the more serious the drift phenomenon deviating from the set value. The system has a built-in two-dimensional lookup table form of the tank reaction coefficient conversion table. The table uses row indices to represent discretized temperature fluctuation index intervals and column indices to represent discretized pH drift index intervals. Each cell stores a preset tank reaction coefficient value. The core design rule of this table is: as the row index temperature fluctuation index increases, the temperature fluctuation intensifies or the column index pH drift index increases, the pH drift intensifies, and the stored tank reaction coefficient value monotonically decreases. For example, the tank reaction coefficient corresponding to the temperature fluctuation index in the 0 to 0.5 interval and the pH drift index in the 0 to 0.01 interval may be 0.95; while the tank reaction coefficient corresponding to the temperature fluctuation index in the 1.0 to 1.5 interval and the pH drift index in the 0.02 to 0.03 interval may decrease to 0.65. The system takes the real-time calculated temperature fluctuation index and pH drift index values as inputs to determine their respective row interval and column interval, and outputs the corresponding tank reaction coefficient value through the lookup table operation. This coefficient represents the comprehensive stability of the current physical and chemical environment in the tank, and the lower the value, the more unstable the environment, and the stronger the potential inhibition of the thickener performance.

[0055] The system receives the current viscosity value measured by the online viscosity sensor in real time and compares it with the target viscosity value preset by the user, calculating the viscosity difference in centipoise. This difference serves as the basic input for rate adjustment requirements. The system calls the preset basic rate conversion relationship, which defines the amount of thickener injection rate adjustment required to eliminate a specific viscosity difference under ideal stable working conditions. This relationship is usually implemented in the form of a function: the initial thickener injection rate :

[0056]

[0057] wherein: is the viscosity difference. The function is determined by process experience and fluid characteristics, for example, a proportional control is used, i.e. the base rate is proportional to the viscosity difference, and the proportional coefficient is preset according to the fluid characteristics. If the viscosity difference is negative, the current viscosity is higher than the target, and the base rate calculation result is negative, indicating that the thickener injection amount needs to be reduced. The system has a built-in coefficient influence weight table. The table includes two parts: the first part relates the tank reaction coefficient to the first adjustment weight. This part usually divides the tank reaction coefficient value into several intervals, such as 0.9 to 1.0 interval, 0.7 to 0.9 interval, etc., and specifies a weight value for each interval. The design principle is: the lower the tank reaction coefficient value, the more unstable the environment, and the weight is usually set to a negative value and the absolute value is larger, for example, the tank reaction coefficient is in the 0.9 to 1.0 interval, the weight is 0, in the 0.7 to 0.9 interval, the weight is -0.1, in the 0.5 to 0.7 interval, the weight is -0.2. The second part relates the stirring state coefficient to the second adjustment weight. Similarly, the stirring state coefficient value is divided into intervals, and a weight value is specified. The design principle is: the lower the stirring state coefficient value, the worse the stirring efficiency, and the weight is usually set to a negative value and the absolute value is larger, for example, the stirring state coefficient is in the 0.9 to 1.0 interval, the weight is 0, in the 0.7 to 0.9 interval, the weight is -0.05, in the 0.5 to 0.7 interval, the weight is -0.15. The system queries the first part of the table according to the current calculated tank reaction coefficient value to obtain the corresponding first adjustment weight, and queries the second part of the table according to the current input stirring state coefficient value to obtain the corresponding second adjustment weight. The system performs weighted correction calculation to generate the final target thickener injection rate . The correction formula is:

[0058]

[0059] wherein: represents the target thickener injection rate, unit: liters / minute, represents the initial thickener injection rate, unit: liters / minute, represents the first adjustment weight, represents the second adjustment weight. The formula embodies the combined correction effect of the tank reaction state and the stirring state on the base rate. For example, assuming that the base rate is calculated to be 10 liters / minute, indicating that the injection amount needs to be increased, the current tank reaction coefficient is 0.8, corresponding to the first adjustment weight -0.1, and the stirring state coefficient is 0.75, corresponding to the second adjustment weight -0.1, then the target rate = 10 x (1-0.1-0.1) = 8 liters / minute. The calculation result is less than the base rate, reflecting the negative impact of unstable environment and low-efficiency stirring on the performance of the thickening agent, and the system therefore reduces the injection rate to avoid invalid addition or local over-thickening. If the environment is stable and the tank reaction coefficient is high, the first adjustment weight is close to 0 and the stirring is efficient, the stirring state coefficient is high, and the second adjustment weight is close to 0, then the target rate is approximately equal to the base rate, and the injection is made according to the base requirement.

[0060] The operation of this embodiment relies on continuous data flow and preset configuration data. The temperature and pH sensor signals are transmitted to the data acquisition card through signal lines, with sampling frequency and accuracy meeting process requirements, such as temperature ±0.1°C and pH ±0.01. The calculation of fluctuation index and drift index is performed periodically, such as once every second, on the latest window data in the embedded processor or industrial control computer. The tank reaction coefficient conversion table is stored in the system configuration file, allowing process engineers to adjust the interval division and coefficient assignment according to the sensitivity of the specific reaction system. The calculation of viscosity difference is also performed periodically, and its sign and size directly drive the calculation direction and amplitude of the base rate. The coefficient influence weight table is a separate data structure, and the assignment of its weight is based on the analysis and induction of the influence degree of environmental factors and stirring factors on thickening efficiency in historical data. The weighted correction formula is the core calculation step of the rate optimization model, and the output target rate is a signed flow setting value, positive for injection and negative for reducing injection. In the entire process, time stamping is crucial to ensure that the data used to calculate temperature fluctuation index, pH drift index, viscosity difference, tank reaction coefficient, and stirring state coefficient correspond to the same or overlapping time period, to ensure the spatiotemporal consistency of state evaluation. The system performs validity checks on input data, such as sensor disconnection, data out of range or jumping, and replaces them with the last valid value or a preset safety value, and triggers an alarm. The intermediate variables and final output target rate in the calculation process are recorded in the operation log with accurate time stamp and equipment identification, meeting the process tracing and fault diagnosis requirements. The operation interface can display these key parameters and their change trends in real time. This embodiment integrates information on tank environment stability, stirring efficiency, and viscosity deviation, dynamically generates a thickening agent injection rate command that adapts to the actual working conditions through table lookup and weighted calculation, and transmits the command to the execution control module to drive physical devices to act.

[0061] Embodiment 3: This embodiment involves a technical solution in which the system monitors the actual viscosity response after the injection rate of thickening agent is adjusted, and dynamically calibrates the internal model parameters based on the monitoring results. The system continuously collects viscosity measurement values of the reaction system at a fixed sampling frequency through a high-precision online viscosity sensor to form a raw data sequence. To quantify the rate of change of viscosity over time, the system performs the following calculation steps: calculate the absolute change in viscosity between adjacent sampling points, i.e., for each sampling point, calculate the absolute value of the difference between adjacent viscosity measurements, generating an absolute change sequence. Moving average processing is performed on this sequence to smooth random noise and extract the trend change rate. The system sets a fixed-length moving window containing a number of absolute change values, for example, 15 points representing 1.5 seconds of data. The formula for calculating the actual viscosity change gradient is:

[0062]

[0063] where: represents the unit of actual viscosity change gradient: centipoise / second, is a preset positive integer of moving window length, is the current latest sampling point index, is the traversal index of the sampling points in the window, is the absolute change in viscosity corresponding to the th sampling point in the window. The summation range is from to , i.e., including the latest absolute value change quantities in the window. This gradient characterizes the average absolute change rate of the system viscosity within the recent time window. The system internally stores a pre-set expected gradient range based on process knowledge, which defines the reasonable interval in which the viscosity response change rate should be after the injection rate of thickening agent is adjusted, for example, the lower limit is 0.5 centipoise / second and the upper limit is 2.0 centipoise / second. The system continuously compares whether the calculated actual gradient falls within the expected range. If the actual gradient is lower than the lower limit, the response is too slow or higher than the upper limit, it is determined that the actual viscosity change gradient deviates from the pre-set expected gradient range, triggering the generation of model calibration instructions. This instruction contains information such as deviation type identification and time stamp.

[0064] ​When the model calibration instruction is generated, the system automatically analyzes the pattern of the deviation cause. If the deviation phenomenon, such as the slow response, is strongly associated with the output of the stirring state model. For example, the historical data shows that the response delay frequently occurs when the stirring state coefficient is low, the instruction will specify to update the torque deviation level table inside the stirring state model. The update operation is performed by an authorized user through a maintenance interface or automatically by the system according to pre-set rules. The specific update strategy can include: adjusting the interval boundary of the absolute value of the torque deviation amount, such as modifying the original 1.0 to 2.5 Newton per meter interval to 0.8 to 2.2 Newton per meter, or adjusting the coefficient value corresponding to the interval, increasing the coefficient value of the original interval from 0.8 to 0.85. If the deviation phenomenon, such as the fast response with oscillation, is strongly associated with the output of the tank reaction model. For example, the over-adjustment frequently occurs when the tank reaction coefficient is low, the instruction will specify to update the tank reaction coefficient conversion table inside the tank reaction model. The update operation is also performed through the interface or automatic rules, and the strategy can include: adjusting the interval division of the temperature fluctuation index or the pH drift index, or adjusting the coefficient value of a specific cell. The updated mapping table will replace the old table stored in the non-volatile memory for subsequent coefficient calculation in the period. This process aims to correct the static mapping relationship inside the model, so that the model output more accurately reflects the influence of the actual working condition on the thickening process.

[0065] The system runs an overshoot monitoring counter independently. The counter monitors every viscosity regulation process (i.e. from the start of the rate adjustment to the viscosity stabilizing near the target value). The system defines a viscosity regulation overshoot event: when the real-time viscosity measurement value first reaches or exceeds the target viscosity, if it continues to rise and exceeds the preset upper limit deviation threshold, it is counted as an overshoot event. At the end of each regulation process, if overshoot occurs, the counter is incremented by 1. The system sets a statistical period, for example, a complete production batch or a continuous 8-hour operation cycle, and continues to accumulate the number of overshoot events within the period. The system presets a threshold value. When the statistical period ends, if the number of overshoots reaches or exceeds the threshold value, a weight adjustment instruction is generated. The instruction contains the statistical value of the number of overshoots and the period of occurrence. The rate optimization module responds to this instruction and analyzes the correlation pattern of overshoot events and weight factor usage. If the analysis shows that overshoot events occur in the period of unstable environment with low tank reaction coefficient, it is determined that the current setting of the first adjustment weight may be too large in absolute value, causing excessive inhibition of the injection rate by the weighted correction, leading to subsequent compensatory overshoot. At this time, the system resets the weight values associated with the tank reaction coefficient interval in the coefficient influence weight reference table, with the strategy of reducing the strength of its negative impact. If the analysis shows that overshoot events occur in the period of poor stirring efficiency with low stirring state coefficient, it is determined that the current setting of the second adjustment weight may be too large. At this time, the system resets the weight values associated with the stirring state coefficient interval in the reference table, also reducing the strength of its negative impact. The resetting of weight values is achieved by modifying the corresponding entries in the configuration file or database. This process aims to optimize the dynamic response characteristics of the control system by adjusting the relative weight influence of the tank reaction state and the stirring state in the final rate decision, and to reduce overshoot oscillation phenomena. The entire feedback calibration process relies on accurate collection of viscosity response data, gradient calculation, event judgment, and parameter update mechanism, forming a closed-loop optimization loop.

[0066] Embodiment 4: refer to Figure 4 The present embodiment covers the environmental factor compensation calculation and safety interlock control mechanism of the thickening agent injection rate. The system receives the target thickening agent injection rate instruction signal , and then starts the multi-physical parameter compensation correction process. The high-precision liquid level meter installed in the thickening agent storage tank collects real-time liquid level data , unit: meter, data update rate after filtering: 2Hz. The pressure transmitter is used to obtain the pressure value of the conveying main pipeline , unit: megapascal, data sampling rate: 4Hz, and moving average filtering is used. In order to accurately represent the coupling effect of the storage tank liquid level and the pipeline pressure on the target flow, the system has an integrated parameter table named liquid pressure working condition compensation matrix. The matrix table uses a three-dimensional index structure, including liquid level height classification interval, pressure difference partition and corresponding flow correction parameters, refer to Table 1.

[0067] Table 1: Integrated parameter table of hydraulic working condition compensation matrix

[0068] Liquid level partition (m) Differential pressure partition ΔP (MPa) Dynamic compensation factor (κ) Flow offset (δ, L / min) [2.8, 3.2] [0.00, 0.03) 1.00 0.00 [2.8, 3.2] [0.03, 0.07) 0.98 +0.15 [2.8, 3.2] [0.07, 0.12) 0.96 +0.30 [2.8, 3.2] [0.12, 0.18) 0.94 +0.45 [2.8, 3.2] [0.18, 0.23) 0.92 +0.60 [2.8, 3.2] ≥0.23 0.90 +0.75 [2.4, 2.8] [0.00, 0.03) 1.02 +0.05 [2.4, 2.8] [0.03, 0.07) 1.00 +0.20 [2.4, 2.8] [0.07, 0.12) 0.98 +0.35 [2.4, 2.8] [0.12, 0.18) 0.96 +0.50 [2.4, 2.8] [0.18, 0.23) 0.94 +0.65 [2.4, 2.8] ≥0.23 0.92 +0.80 [2.0, 2.4] [0.00, 0.03) 1.05 +0.10 [2.0, 2.4] [0.03, 0.07) 1.03 +0.25 [2.0, 2.4] [0.07, 0.12) 1.01 +0.40 [2.0, 2.4] [0.12, 0.18) 0.99 +0.55 [2.0, 2.4] [0.18, 0.23) 0.97 +0.70 [2.0, 2.4] ≥0.23 0.95 +0.85 [1.2, 1.6] [0.00, 0.03) 1.10 +0.25 [1.2, 1.6] [0.03, 0.07) 1.08 +0.40 [1.2, 1.6] [0.07, 0.12) 1.06 +0.55 [1.2, 1.6] [0.12, 0.18) 1.04 +0.70 [1.2, 1.6] [0.18, 0.23) 1.02 +0.85 [1.2, 1.6] ≥0.23 1.00 +1.00 [0.4, 0.8] [0.00, 0.03) 1.15 +0.45 [0.4, 0.8] [0.03, 0.07) 1.13 +0.60 [0.4, 0.8] [0.07, 0.12) 1.11 +0.75 [0.4, 0.8] [0.12, 0.18) 1.09 +0.90 [0.4, 0.8] [0.18, 0.23) 1.07 +1.05 [0.4, 0.8] ≥0.23 1.05 +1.20

[0069] The table covers 5 liquid level partitions (2.8-3.2m, 2.4-2.8m, 2.0-2.4m, 1.2-1.6m, 0.4-0.8m) and 6 pressure difference partitions (0.00-0.03MPa, 0.03-0.07MPa, 0.07-0.12MPa, 0.12-0.18MPa, 0.18-0.23MPa, ≥0.23MPa), forming 30 groups of compensation parameter combinations. The system presets the standard working condition pressure reference value, calculates the absolute value of the difference between the actual pressure and the reference pressure. Match the current effective liquid level height with the liquid level partition, match the pressure difference value with the pressure difference partition, and obtain the compensation factor and offset through matrix cross query. Finally, the execution rate is reconstructed according to the following relationship:

[0070] Operation execution rate (

[0071] For example, the liquid level height is 1.35 meters, which belongs to the 1.2-1.6 meter partition, the actual pressure is 0.25 MPa, and the pressure difference is 0.17 MPa, which belongs to the 0.12-0.18 MPa partition, the compensation factor is 1.04 and the offset is 0.70 liters / minute according to the table. If the target rate is 12.5 liters / minute, the operation execution rate is (12.5 x 1.04) + 0.70 = 13.7 liters / minute. The compensation calculation is completed in 50 milliseconds, and the output resolution is 0.01 liters / minute. The historical data recording module synchronously stores the input parameters, query parameters and output results to form a compensation process tracking chain. The system establishes an independent safety interlock monitoring network, which directly connects the safety instrument system through a double-redundant safety bus. The emergency state signals monitored include three types: type A reaction tank body temperature overrun: activated when the temperature detection value exceeds 85.0°C, type B reaction pressure overrun: activated when the pressure exceeds 0.78 MPa, and type C stirring shaft radial vibration overrun: activated when the vibration value exceeds 7.1 mm / s. The system has built-in safety response logic rules that define the disposal strategies under different emergency state combinations. When a single type A signal is detected, the close injection valve action is performed, the forced zero rate override mode is adopted, and the three-level alarm is triggered; when a type B signal is detected, the pump stop and pressure relief action is performed, the forced zero mode is adopted, and the two-level alarm is triggered; when a type C signal is detected, the speed reduction to 50% operation action is performed, the proportional limiting mode is adopted, and the one-level alarm is triggered. When type A and type B signals are detected at the same time, the emergency stop action sequence is performed, the forced zero mode is adopted, and the special-level alarm is triggered; when type B and type C signals are detected, the pump stop and standby pipeline switching action is performed, the instruction transfer mode is adopted, and the special-level alarm is triggered; when type A and type C signals are detected, the close injection valve and vibration suppression action is started, the forced zero mode is adopted, and the two-level alarm is triggered.

[0072] When any emergency signal is detected, the system completes signal type recognition and logic matching within 12 milliseconds. According to the matching result, the corresponding execution instruction is generated, for example, when type A and type B signals are detected at the same time, the emergency stop instruction sequence is triggered: send the emergency stop code to the injection device through the safety bus; activate the reaction tank protection program; cover the rate output channel to zero. The execution state of all operation instructions is verified through hardwired feedback to form a closed loop control, and the single control cycle consumes no more than 80 milliseconds. During the emergency disposal period, the system records signal state, execution instruction code, device feedback signal and compensation calculation state at a frequency of 20 frames per second to form an encrypted event stream data packet.

[0073] The emergency state is lifted under three conditions: all trigger signals return to the safe range for 30 seconds; the tank temperature drops to the safe threshold interval; and the operating table is manually reset. The system then starts a five-stage recovery protocol: the first stage initializes sensor verification and removes the rate coverage lock; the second stage reacquires the latest target rate value; the third stage performs current hydraulic working condition compensation calculation; the fourth stage implements the execution rate in a gradient incremental manner; and the fifth stage switches to normal operation mode. The system automatically generates a recovery process report, including timestamps for each stage, compensation parameter change curve, rate climbing gradient, and final stable deviation value. The safety log system uses a block storage mechanism: real-time operation data are stored in cache for 7 days; event stream data packets are stored in encrypted solid state disks for 6 months; and audit tracking data are written to read-only media for 10 years. The configuration table maintenance management adopts a multi-level authorization principle: process engineers propose parameter modifications, device engineers confirm hardware compatibility, safety engineers assess risks, and system administrators authorize execution with double passwords. Each configuration change generates an unalterable record containing the old value, new value, modifier digital signature, and timestamp, which is pushed to a third-party audit platform. The system automatically performs compensation matrix validity verification every month: by simulating the injection device to apply a standard flow step signal, collecting the actual output curve, and comparing it with the theoretical compensation value, a parameter calibration alert is triggered if the deviation exceeds the allowed threshold.

[0074] Embodiment 5: This embodiment details the construction method of the reaction tank connection relationship in the production overview and the definition of related parameters. Users set up all container units involved in the post-adjustment stage and their physical connection methods through a graphical configuration interface or structured data import function. Container units include main reaction tanks identified as R-101, intermediate buffer tanks identified as B-201, finished product temporary storage tanks, etc., each tank having a unique device identifier. Connection relationship refers to the material transfer path between tanks, which needs to be clearly defined as the communication between the outlet of the upstream tank and the inlet of the downstream tank. The system requires users to specify all pipeline segments on the connection path, identified as PL-401, conveying equipment centrifugal pump P-301, control valve regulating valve V-501, and their connection order. For example, users configure “main reaction tank R-101 bottom outlet outlet valve XV-101 conveying pump P-301 inlet pump outlet pipeline PL-401 buffer tank B-201 top inlet”. This configuration completely describes the physical path of material flow from the main reaction tank to the buffer tank, referred to as the fluid communication path. The system analyzes user input and constructs a machine-readable topological network structure, where tanks are nodes and pipelines, valves, pumps, etc. are directed edges, with edge direction representing material flow direction. This topological structure is stored in the system database as the basic framework of the global process layout.

[0075] In the defined fluid communication path, the system requires the user to specify one or more thickener injection points. Each injection point needs to be precisely located in its specific position in the topological path. The position definition contains a physical attribution description and a topological position description. The physical attribution description indicates on which specific equipment or pipe section the injection point is located, such as "installed on pipe PL-401, 1.5 meters downstream from the inlet flange of buffer tank B-201". The topological position description emphasizes its relative position relationship in the overall path, such as "located on the connecting pipeline between the main reaction tank R-101 and the buffer tank B-201, downstream of the delivery pump P-301, upstream of the buffer tank B-201". The system assigns a unique identifier to each injection point, such as InjPoint-01, and stores its topological position information, such as predecessor node, successor node, edge identifier where it is located, and physical location coordinates. At the same time, the user needs to configure the initial parameters of the injection point, such as the allowed injection rate range , the associated thickener type code , the maximum allowed injection pressure , etc. These parameters are stored in association with the injection point identifier. This step precisely defines the precise location of the additive introduction and its context environment in the fluid path.

[0076] The system associates monitoring and control resources to key nodes in the production overview. Key nodes include: major material inlets and outlets of each tank, such as R-101 outlet, B-201 inlet, key intermediate points on fluid communication paths, such as PL-401 mid-pipe, and all thickener injection points, such as InjPoint-01. For each key node, the system requires the user to bind a viscosity monitor device identifier for online monitoring of fluid viscosity at the node. The identifier points to a specific sensor device, such as "VISSensorR101Outlet" for a viscometer installed at the main reactor outlet pipe, "VISSensorB201Inlet" for a buffer tank inlet viscometer, and "VISSensorPL401Mid" for a mid-pipe viscometer. These identifiers serve as addressing basis for the system to collect real-time viscosity data u(t). Similarly, for each thickener injection point, the system requires the user to bind a thickener valve control identifier for controlling the thickener flow rate. The identifier points to a specific actuator, such as "FCVInj01" for a regulating valve controlling the flow rate at injection point InjPoint-01, and "PCV_InjTank" for a thickener tank outlet pressure control valve. These identifiers serve as target addresses for the system to issue injection rate control commands u(t). The system establishes and stores a mapping table between node locations (topological and physical locations), monitor device identifiers, and control valve identifiers. The mapping enables the system to accurately associate sensor data to specific locations based on the topological network, and to precisely issue control commands to target actuators. The entire production overview construction process eventually generates a structured dataset containing vessel connection topology, injection point localization, and device identifier mapping, providing a global path map and device address book for data collection, state assessment, optimization decision, and control execution.

[0077] It is to be understood that the terms such as first and second, etc., are used herein merely to distinguish one general concept from another, and do not necessarily require or imply any such actual relationship or order between or among such entities or operations. Also, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0078] While the embodiments of the present application have been illustrated and described, it will be understood by those skilled in the art that various changes, modifications, substitutions, and alterations can be made therein without departing from the principles and spirit of the application, and it is intended that the scope of the application be limited solely by the appended claims and their equivalents.

Claims

1. A system for adding thickener and controlling viscosity online in the post-adjustment stage of polyacrylic acid emulsion, characterized in that, include: The overview building module is used to create a production overview based on user settings. The production overview includes the connection relationships of the reaction tanks and the parameters of the thickener injection nodes. The coefficient generation module is used to collect real-time stirring state data and tank reaction data, construct a stirring state model based on the stirring state data and output stirring state coefficients, and construct a tank reaction model based on the tank reaction data and output tank reaction coefficients. The rate optimization module is used to obtain the viscosity difference between the current viscosity measurement value and the target viscosity, and to construct a rate optimization model based on the tank reaction coefficient, the stirring state coefficient and the viscosity difference, and output the target thickener injection rate. The execution control module is used to drive the thickener injection device according to the target thickener injection rate; The coefficient generation module constructs a stirring state model based on the stirring state data and outputs stirring state coefficients, including: retrieving a preset mapping relationship between stirring speed and blade torque, and matching the current stirring speed data to obtain the theoretical torque reference value; Calculate the torque deviation between the actual measured torque value and the theoretical torque reference value; The stirring state coefficient is determined based on the torque deviation and a preset torque deviation level table, wherein the stirring state coefficient decreases accordingly when the torque deviation increases. The coefficient generation module constructs a tank reaction model based on the tank reaction data and outputs the tank reaction coefficients, including: acquiring reaction temperature time series data and pH value time series data; The temperature fluctuation index is obtained by extracting fluctuation features from the reaction temperature time series data, and the pH drift index is obtained by extracting drift features from the pH value time series data. Input the temperature fluctuation index and the pH drift index into a preset tank reaction coefficient conversion table, and output the tank reaction coefficient accordingly. The rate optimization module constructs a rate optimization model based on the tank reaction coefficient, the stirring state coefficient, and the viscosity difference, including: calling a preset coefficient influence weight comparison table, determining a first adjustment weight based on the tank reaction coefficient, and determining a second adjustment weight based on the stirring state coefficient; The initial thickener injection rate is calculated based on the viscosity difference and the preset basic rate conversion relationship. The initial thickener injection rate is weighted and corrected by combining the first adjustment weight and the second adjustment weight to generate the target thickener injection rate.

2. The thickener addition and online viscosity control system for the post-adjustment stage of polyacrylic acid emulsion according to claim 1, characterized in that, The system also includes a feedback calibration module; the feedback calibration module is used to collect the actual viscosity change gradient after executing the target thickener injection rate; When the actual viscosity change gradient deviates from the preset expected gradient range, a model calibration command is generated; The coefficient generation module responds to the model calibration command by updating the torque deviation level table in the stirring state model or updating the tank reaction coefficient conversion table in the tank reaction model.

3. The thickener addition and online viscosity control system for the post-adjustment stage of polyacrylic acid emulsion according to claim 2, characterized in that, The feedback calibration module is also used to count the number of viscosity control overshoots within a preset time period. When the viscosity control overshoot count reaches a preset threshold, a weight adjustment instruction is generated; The rate optimization module responds to the weight adjustment instruction by resetting the first adjustment weight or the second adjustment weight in the coefficient influence weight lookup table.

4. The thickener addition and online viscosity control system for the post-adjustment stage of polyacrylic acid emulsion according to claim 1, characterized in that, The execution control module drives the thickener injection device according to the target thickener injection rate, including: acquiring the current liquid level height and pipeline pressure data of the thickener storage tank; and calculating the injection rate compensation amount according to the current liquid level height and the preset liquid level compensation relationship. The injection rate attenuation is calculated based on the pipeline pressure data and the preset pressure attenuation relationship. The target thickener injection rate is superimposed with the injection rate compensation amount and the injection rate attenuation amount is subtracted to generate the final execution rate, which is then output to the thickener injection device.

5. The thickener addition and online viscosity control system for the post-adjustment stage of polyacrylic acid emulsion according to claim 4, characterized in that, The system also includes a priority processing module; the priority processing module is used to monitor emergency status signals of the reaction vessel. When a temperature or pressure over-limit signal is received, an injection interrupt command is generated to override the final execution rate. When the emergency signal is lifted, the output of the final execution rate is restored.

6. The thickener addition and online viscosity control system for the post-adjustment stage of polyacrylic acid emulsion according to claim 2, characterized in that, When the feedback calibration module acquires the actual viscosity change gradient, it includes: recording the continuous sampling values ​​of the viscosity sensor within a unit time. Calculate the sequence of absolute changes between adjacent sample values; The actual viscosity change gradient is obtained by performing a moving average on the absolute change sequence.

7. The thickener addition and online viscosity control system for the post-adjustment stage of polyacrylic acid emulsion according to claim 1, characterized in that, The production overview created by the overview building module includes the connection relationships of the reaction tanks, including: identifying the fluid communication path between the main reaction tank and the buffer tank; Define the topological location of the thickener injection point in the fluid communication path; Associate the viscosity monitoring device identifiers and thickener valve control identifiers for each node.

Images