Modularized intelligent mine wastewater treatment system

Abstract

The invention relates to the technical field of water treatment, and discloses a modular intelligent mine wastewater treatment system, which comprises a data acquisition module, a data processing module and a control module, the calculation module is used for calculating a zero delay reaction rate constant by utilizing Smith estimation; the decision control module is used for determining a virtual topological structure and driving a valve matrix; the configuration module is used for mapping the physical unit to a logic position based on the chemical saturation index; and the dosing execution module is used for executing gradient collaborative dosing according to the stage number. The real-time apparent reaction rate constant with the zero-delay characteristic is solved by introducing the Smith estimation compensation algorithm, constructing the nominal model and superposing the actual observation deviation, the real working condition can be obtained at the moment of pollutant concentration fluctuation or reaction activity change, and the real-time performance of the real-time apparent reaction rate constant is improved. Therefore, the topology reconstruction decision is completed before the degradation-resistant wastewater flows through the main treatment unit, and the phenomenon that initial discharge exceeds the standard due to insufficient timeliness of traditional feedback control is avoided.

Description

Technical Field

[0001] This invention relates to the field of water treatment technology, and in particular to a modular intelligent mine wastewater treatment system. Background Technology

[0002] Mine wastewater is characterized by large discharge volumes, high levels of suspended solids, and drastic fluctuations in water quality. Its treatment typically relies on the synergistic effect of physical sedimentation and chemical reactions. Existing mine wastewater treatment facilities mostly employ fixed-volume reaction tanks in civil engineering form, or several standardized treatment modules connected in a predetermined sequence. While this static physical structure design meets the basic design requirements, it struggles to adapt to the nonlinear dynamic changes in mine inflow and pollutant concentration. When the influent water quality suddenly deteriorates, the fixed reaction units cannot temporarily expand the physical flow to extend the hydraulic retention time, resulting in incomplete reactions. Conversely, when the influent load is low, the fixed multi-stage series flow limits the overall hydraulic flux of the system, leading to low equipment utilization and energy waste.

[0003] At the process control level, existing systems generally rely on feedback control loops based on effluent quality monitoring. However, coagulation, sedimentation, and redox reactions in wastewater treatment all exhibit significant physical inertia and large time delays. Fluid must undergo a fixed hydraulic residence time from the inlet to the effluent monitoring point. Traditional feedback mechanisms ignore this pure time delay, causing the control system's adjustments to always lag behind actual changes in influent quality. When high-concentration shock loads enter the system, the controller often only responds after the lag time has ended and effluent indicators have already exceeded limits. This mismatch between the control cycle and the reaction cycle makes the system highly susceptible to instantaneous discharge exceeding standards within the initial response blind zone when dealing with sudden pollution loads.

[0004] Furthermore, in scenarios involving the coordinated operation of multiple treatment units, existing scheduling strategies often lack the identification and utilization of the microscopic chemical state of each unit. Traditional control methods typically treat all isomorphic units as uniform black boxes, employing random allocation or simple sequential polling strategies during start-up, shutdown, or process switching, ignoring the differences in activated sludge concentration or reagent residue accumulated by each unit due to different historical operating conditions. This indiscriminate scheduling method easily leads to uneven heating and cooling: for example, suddenly placing a clean standby unit at the high-load head end can easily trigger chemical breakthrough; or placing a unit with a large accumulation of reaction byproducts at the tail end can lead to secondary release of pollution. At the same time, the lack of inter-stage coordinated dosing methods often results in insufficient dosing at the head end or excessive dosing at the tail end, and the recovery and utilization of residual reagents in the effluent are not considered, leading to severe chemical field oscillations during system condition switching and persistently high reagent costs in the long term. Summary of the Invention

[0005] The purpose of this invention is to provide a modular intelligent mine wastewater treatment system, which solves the problems of existing wastewater treatment systems being unable to dynamically adapt to fluctuations in water quality and quantity, and the traditional feedback control system causing response delays due to lag, as well as the oscillations caused by switching and excessive consumption of chemicals.

[0006] To achieve the above objectives, the present invention provides a modular intelligent mine wastewater treatment system, including a data acquisition module, a calculation module, a decision control module, a configuration module, and a dosing execution module.

[0007] The data acquisition module is used to digitally map each wastewater treatment pond in the mine wastewater treatment system, constructing a set of physically existing wastewater treatment ponds into a reaction unit set.

[0008] The central controller designates one reaction unit in the reaction unit set as a pilot unit for monitoring, and collects in real-time influent and effluent pollutant concentrations from all units with time synchronization. During the operation of the pilot unit, the system sets its flow rate to the optimal design flow rate for the individual unit to eliminate the interference of flow fluctuations on the measurement of kinetic parameters; simultaneously, the pilot unit's dosing system is configured in proportional flow following mode to ensure a constant ratio of reagent dosage to influent flow rate.

[0009] The computation module overcomes the hydraulic residence time lag in the physical system by executing the Smith predictor compensation algorithm to construct an instantaneous mapping from monitoring data to the actual reaction state, thus calculating the real-time apparent reaction rate constant with zero latency. Specifically, the computation module first constructs a nominal model without pure time delay, based on the influent pollutant concentration collected at the current moment. Calculate the lag-free prediction output :

[0010] ;

[0011] In the formula, The preset baseline reaction rate constant, The effective reaction volume of the pilot unit, The pilot unit's operating flow rate is used. Simultaneously, a model incorporating hysteresis is employed, based on the hydraulic residence time. Previous influent concentration data calculation lag prediction output and the actual effluent concentration monitoring value The model prediction bias is obtained by performing a difference operation with the lagged prediction output. :

[0012] ;

[0013] Subsequently, the model prediction bias is superimposed onto the hysteresis-free prediction output to generate a synthetic feedback signal. The real-time apparent reaction rate constant is obtained by reverse calculation using this signal. :

[0014] ;

[0015] The decision control module is used to base decisions on the real-time apparent reaction rate constant. and the preset target effluent concentration The module reverse-engineers the physical processing capacity required by the unit and determines the virtual cluster topology. It first calculates the theoretical total reaction time required to degrade the current influent pollutant concentration to the target effluent concentration. And introduce an engineering safety factor :

[0016] ;

[0017] Next, the theoretical total reaction time is divided by the hydraulic residence time of a single unit. Then round up to obtain the minimum number of cascade stages required for a single processing link. Then, the total number of actual usable units after deducting the pilot unit is calculated. Divide by the minimum number of cascaded stages and round down to obtain the maximum number of parallel chains. :

[0018] ;

[0019] ;

[0020] The decision control module generates hardware driving instructions based on the calculation results to control the actions of the reconfigurable valve matrix: switching the water outlet valves of adjacent units in the virtual cluster topology to the water inlet of the next unit to form an inter-stage series path; connecting the water inlet valves of the first unit of each link to the system's main water inlet distribution pipe; and controlling the lateral interconnection valves between different parallel links to close, ensuring that each parallel link is hydraulically independent.

[0021] The configuration module is used to establish and update the chemical saturation index of each cell in the cell set in real time. The chemical saturation index is used to map physical units to logical locations within the virtual cluster topology. The chemical saturation index is defined as the chemical saturation index of a unit within a specific time sliding window. The cumulative integral of the chemical reaction load borne by the unit is recorded in real time by the concentration of influent pollutants. and real-time dosage The product of the two, after time weighting, is integrated to obtain:

[0022] ;

[0023] During mapping and allocation, the configuration module sorts all available units in descending order according to their chemical saturation index; selects the units with higher rankings to map to the first-level position of the virtual cluster topology to utilize their chemical inertia to respond to the influent load; selects the units with lower rankings to map to the last-level position of the virtual cluster topology to utilize their low residual environment to prevent secondary pollution; and fills the remaining units into the intermediate-level positions.

[0024] The dosing execution module is used to implement differentiated gradient-coordinated dosing strategies for reaction units at different levels within the virtual cluster topology. For units mapped to the first-level position, an enhanced feedforward control strategy is executed: the model prediction bias generated by the computation module is invoked. As a correction factor, based on the basic dosage coefficient Current operating flow rate Real-time influent pollutant concentration Calculate the dosage :

[0025] ;

[0026] For cells mapped to intermediate levels, a proportional transfer control strategy is implemented: the dosage of the current-level cell is set as a decay function of the dosage of the previous-level cell, and the inter-stage transfer decay coefficient is dynamically adjusted based on the real-time apparent reaction rate constant. :

[0027] ;

[0028] For units mapped to the final stage, a complementary reverse feedback control strategy is implemented: a reconfigurable valve matrix is ​​driven to construct a reverse flow path connecting the outlet of the final stage unit to the mixing zone at the system's inlet; a PID closed-loop feedback algorithm including a backflow reagent deduction term is used, based on the deviation between the measured concentration of the final stage effluent and the safety control target value. Perform proportional-integral-differential calculations, and then the central controller uses this deviation as a basis for further analysis. Perform proportional-integral-differential (PID) calculations, and subtract the equivalent amount of reagent saved due to the reflux effect. Therefore, the final dosage calculation formula is as follows:

[0029] ;

[0030] In the formula, , , These are the proportional, integral, and differential gain coefficients, respectively. This represents the current reverse flow traffic; The influent flow rate of the primary unit; This refers to the current average dosage level of the entire wastewater treatment system, that is, all units. The conversion factor for the activity of refluxed reagents ( ).

[0031] In summary, the present invention has at least one of the following beneficial technical effects:

[0032] 1. This invention introduces the Smith prediction compensation algorithm, constructs a lag-free nominal model and superimposes actual observation biases to calculate a real-time apparent reaction rate constant with zero delay characteristics. It can obtain the real operating conditions at the moment when pollutant concentration fluctuates or reactivity changes, thereby completing the topology reconstruction decision before the recalcitrant wastewater flows through the main treatment unit, avoiding the phenomenon of initial emission exceeding the standard due to insufficient timeliness of traditional feedback control.

[0033] 2. This invention establishes a chemical saturation index and maps high-saturation units to the first-stage unit to utilize their chemical inertia to resist shock. Therefore, low-saturation units can be mapped to the final stage to prevent secondary pollution. Combined with feedforward correction in the first stage, proportional decay in the intermediate stage, and feedback control with reflux deduction in the final stage, this invention achieves a state-based physical mapping and graded gradient dosing strategy. This not only eliminates chemical oscillations during unit switching but also fully utilizes the remaining reagent activity through reverse reflux, thereby reducing operating costs.

[0034] 3. This invention, through virtual cluster computing logic based on compliance constraints, can automatically adjust the series and parallel connection mode of reaction units according to real-time water quality conditions. When the influent load is high, it automatically increases the number of series stages to extend the effective reaction time, and when the influent load is low, it automatically increases the number of parallel links to increase the total processing throughput. Thus, it can achieve a balance between reaction time and hydraulic load while ensuring stable effluent compliance. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the system architecture of the present invention;

[0036] Figure 2 This is a schematic diagram illustrating the principle of real-time reaction kinetic parameter calculation based on Smith prediction compensation according to the present invention.

[0037] Figure 3 This is a block diagram of the differentiated gradient synergistic dosing control strategy of the present invention;

[0038] Figure 4 This is a schematic diagram illustrating the principle of calculating the chemical saturation index of the present invention;

[0039] Figure 5 This is a comparison curve of the effluent response of the system of the present invention and the traditional fixed process under influent fluctuations. Detailed Implementation

[0040] The following is in conjunction with the appendix Figure 1 -Appendix Figure 5 The present invention will be further described in detail below.

[0041] This invention provides a modular intelligent mine wastewater treatment system, the system comprising:

[0042] The data acquisition module is used to digitally map the treatment pool and set up pilot units, and to collect the concentration of pollutants in the influent and effluent of each unit in real time.

[0043] The computation module is used to execute the Smith prediction compensation algorithm, construct an instant mapping, and solve for the real-time apparent reaction rate constant with zero latency characteristics.

[0044] The decision control module is used to calculate the physical processing capacity based on the reaction rate constant, determine the virtual cluster topology, and drive the reconfigurable valve matrix to operate.

[0045] The configuration module is used to perform mapping and allocation from physical units to virtual topological logical locations based on the real-time updated chemical saturation index.

[0046] The dosing execution module is used to execute differentiated gradient coordinated dosing strategies for different levels of units.

[0047] In one specific embodiment: First, each wastewater treatment tank of the mine wastewater treatment system is digitally mapped. Specifically, the total number is... Several wastewater treatment ponds with identical physical structures are constructed into a unit set. At this time, the central controller will control the number... The reaction unit is configured as a pilot unit for monitoring. The central controller then drives the unit's inlet pump to operate at a frequency that maintains a constant reference flow rate throughout the operating cycle. And this flow rate The optimal design flow rate for a single unit is set to eliminate the interference of flow fluctuations on the measurement of kinetic parameters. Simultaneously, the pilot unit's dosing system is configured in proportional flow following mode to ensure a constant ratio of reagent dosage to influent flow rate. Under this condition, online water quality sensors installed at the inlet and outlet of the pilot unit are used to collect real-time, time-synchronized influent pollutant concentrations. and the concentration of pollutants in the effluent .

[0048] Then, based on the ideal mixing assumption of a continuous stirred reactor, the reaction process within the pilot unit will follow first-order reaction kinetics. However, in actual physical systems, there is inevitably a pure time lag due to hydraulic residence time. Therefore, let the effective reaction volume of the pilot unit be... Then its hydraulic residence time The formula for determining is:

[0049] ;

[0050] in, This represents the physical time required for fluid to flow through the unit, and also indicates the time delay of the effluent monitoring data relative to the influent operating conditions. To describe this physical process mathematically, this implementation constructs a transfer function model that incorporates the reaction mechanism and hysteresis characteristics. Specifically, the transfer function for the wastewater treatment process is defined as follows: From the reaction kinetics part With pure delay It is constructed in series. Among them:

[0051] Reaction kinetics section This reflects the decay process of pollutant concentration under ideal mixing conditions, and according to the first-order reaction kinetic equation, the time-domain differential equation can be described as:

[0052] ;

[0053] In the formula, The instantaneous pollutant concentration within the unit changes over time. The instantaneous pollutant concentration within the unit. The apparent reaction rate constant is used to characterize the current water quality properties. Then, a Laplace transform is performed on the above differential equation to obtain the transfer function form of the reaction process in the complex frequency domain. However, considering the reaction rate constant... In a short time, it can be approximated as a constant; therefore, the reaction kinetics part... Represented as:

[0054] ;

[0055] Taking into account the hydraulic residence time The influence of the global transfer function model of the pilot unit Establish as follows:

[0056] ;

[0057] In the formula, For process gain, The process time constant; For complex variables, use the Laplace transform. The term is a lag term; therefore, this model can accurately describe the complete dynamic process of influent concentration change through reaction decay and time delay, ultimately manifesting as effluent concentration change. Thus, the central controller can combine this lag-based transfer function model with the subsequent Smith prediction method to achieve time-delayed calculation of the actual reaction parameters.

[0058] In one embodiment, after establishing the basic transfer function model that includes hysteresis characteristics, it is also necessary to eliminate the timeliness effect of hydraulic residence time on the solution of reaction parameters. Therefore, this embodiment uses the Smith prediction compensation algorithm executed by the central controller to construct an instant mapping from monitoring data to the actual reaction state.

[0059] Specifically, the algorithm constructs a nominal model without pure time delay to simulate the response output under ideal conditions: the central controller has a pre-set time-delay-free nominal model. This model is based on the reaction kinetics section above. The construction was completed, but the lag factor was removed over time. Furthermore, the nominal model uses the average reaction rate constant statistically derived from the historical data of the unit operation. This serves as a baseline parameter. Then, based on the influent pollutant concentration collected at the current moment... The central controller uses this nominal model to calculate the hysteresis-free predictive output. This predicted value represents the theoretical effluent concentration that the pilot unit should achieve, assuming the fluid reaction is instantaneous and there is no physical transport delay. The calculation is based on the following formula:

[0060] ;

[0061] In the formula, The preset baseline reaction rate constant, The effective reaction volume of the pilot unit, The pilot unit's operating flow rate.

[0062] At the same time, the central controller uses a model with hysteresis characteristics to calculate the hysteresis prediction output. This output value corresponds to the theoretical value of the effluent concentration that should be observed in the actual physical process, and its input uses... The influent concentration data up to a certain time is calculated using the following formula:

[0063] ;

[0064] in, The hydraulic residence time; for The concentration of pollutants in the influent at any given time.

[0065] In order to quantify the difference between the actual reaction process and the model baseline, the central controller can also collect the actual effluent concentration monitoring values ​​of the pilot unit. and compare it with the lag prediction output. Perform difference operations to obtain the model prediction bias. Its formula is expressed as:

[0066] ;

[0067] And this deviation It can not only reflect the actual reaction rate constant relative to the benchmark value Fluctuations can also contain information about unmodeled external disturbances: because and They are aligned in the time dimension, so this deviation can eliminate the interference of lag factors and thus reflect the essential changes in reaction characteristics.

[0068] Based on this, the model prediction bias is then... Superimposed on hysteresis-free prediction output Above, a synthetic feedback signal is generated. And this synthetic feedback signal This refers to the virtual real-time feedback quantity, which includes the current water inflow. Instantaneous change information (from) ), and also combined with actual reaction ability deviation information (from This signal is synchronized with the current water inflow on the time axis, enabling zero-delay characterization of the reaction state.

[0069] Subsequently, the aforementioned synthetic feedback signal was used The central controller performs inverse calculations of the reaction kinetic parameters to obtain the true apparent reaction rate constant at the current moment. Then, based on the principles of first-order reaction kinetics, ... Considered as the current concentration of pollutants in the influent The result after an immediate reaction is... The calculation formula is as follows:

[0070] ;

[0071] Therefore, through this reverse calculation process, the present invention can transform the originally lagging monitoring data into real-time dynamic parameters. And these parameters... It can directly reflect the ease or difficulty of chemical reaction of the wastewater entering the unit, providing forward-looking data for subsequent topology reconstruction and gradient dosing of the virtual cluster, and solving the response lag problem caused by large time delay in traditional feedback control.

[0072] Furthermore, the real-time apparent reaction rate constant with zero delay characteristics was obtained. Subsequently, the central controller enters the decision-making process, specifically calculating the required physical processing capacity of the unit based on the emission standards. Specifically, this implementation first performs a theoretical hydraulic retention time calculation based on compliance constraints, converting the chemical reaction requirements into physical time parameters.

[0073] The fundamental control objective is to ensure that the final effluent pollutant concentration is below the preset emission limit. To this end, the central controller can preset and store the target effluent concentration. This value can be set according to environmental regulations or subsequent process requirements. Therefore, it is based on the current concentration of pollutants entering the system from the influent. and the reaction rate constant calculated in real time At this point, the central controller uses a modified form of the first-order reaction kinetic equation to calculate the theoretical reaction time required for the wastewater to meet discharge standards. However, considering the non-ideal mixing characteristics of the reactor and the potential risks of influent water quality fluctuations in actual engineering projects, a dimensionless engineering safety factor can be introduced into the calculation process. (Typically, the value ranges from 1.1 to 1.3) to provide the necessary redundancy margin. Therefore, the required total reaction time... The specific calculation formula is as follows:

[0074] ;

[0075] In the formula, For the current moment The concentration of pollutants in the influent, The preset target effluent concentration, The real-time reaction rate constant is the one compensated for by Smith's prediction.

[0076] This calculation process essentially establishes a dynamic mapping relationship, specifically: when the reactivity of the wastewater decreases (i.e., (Reduce) or influent concentration When the temperature rises, the calculated required time is... This will be correspondingly extended, indicating the need to construct a longer physical reaction path; conversely, when the influent water quality is good or easily treated, Shortening the parameter indicates a possible shift to a high-throughput operating mode. Therefore, through parameters... It can serve as a quantitative benchmark for subsequent virtual cluster topology reconstruction and physical unit serial series division.

[0077] Subsequently, based on the determined total reaction time requirement, the central controller further performs the topology calculation of the virtual processing chain, discretizing the continuous time parameters into specific unit series levels and parallel links, specifically realizing the mapping from chemical kinetic requirements to physical hardware configuration, ensuring that the processing throughput is maximized while meeting emission standards.

[0078] For a single reaction unit, its physical hydraulic residence time As an inherent property, it is determined by the effective volume of the unit. With the set operating flow rate The ratio is determined. And in order to meet the total reaction time... The requirement is that a single processing link must contain a sufficient number of cascaded units. In this case, a minimum number of cascaded stages can be defined. The minimum number of integer units required to meet the total time requirement. This calculation uses a rounding-up strategy to ensure that the total dwell time of the physical system is not less than the theoretical requirement. The formula is as follows:

[0079] ;

[0080] In the formula, This represents the floor function. Therefore, this formula can be used to determine when the influent water quality deteriorates and leads to... When increasing, As the value increases, longer series links are automatically constructed to extend the response time. After determining the required number of series stages for a single link, the central controller divides the parallel links based on the currently available unit resources in the system. Let the total number of units in the wastewater treatment system be... Subtracting the leading unit as the sensing unit (Usually 1) and the number of units currently under maintenance or in a fault state. This will give you the total number of actually usable units. At this point, with the goal of maximizing processing power, the maximum number of parallel chains that can be constructed is calculated. :

[0081] ;

[0082] In the formula, This represents the floor function. Therefore, this algorithm allows for an automatic optimization mechanism: specifically, increasing the number of cascade stages is necessary when the water quality is poor. At that time, the number of parallel chains Accordingly, the total treatment flow rate is reduced to ensure compliance; however, when the water quality improves, Decrease, automatically increase This increases the overall processing throughput. For units that calculate the remainder, the system sets them to standby mode or assigns them to auxiliary processing stages.

[0083] After completing the virtual topology calculation, the central controller generates corresponding hardware driver instructions to... The cluster structure is transformed into a physical-level valve state matrix. Specifically, this is achieved by defining a valve state matrix. This matrix can contain the opening and closing commands of all interconnected valves between units. The central controller can then execute the following logical mapping: for any constructed virtual link... , and its internal first Level unit ( Switch the outlet valve of ) to the first The inlet end of the first-stage unit forms an inter-stage series path; the inlet valve of the first-stage unit is connected to the main inlet distribution pipe of the system, and the inlet valve of the second-stage unit is connected to the main inlet distribution pipe of the system. The outlet valve of each unit is connected to the main discharge pipe or return pipe. For different parallel links... and In between, the central controller commands the lateral interconnect valves to close, ensuring hydraulic independence for each parallel chain. This is achieved by executing the state matrix. The physical pipeline can be reconfigured in real time to a cluster topology that matches the current water quality characteristics.

[0084] Before achieving physical reconstruction through valve matrix actions, this invention needs to solve a key configuration problem: how to... Each physical unit, with its distinct physical state, is assigned to a specific location within the virtual link. To this end, this implementation introduces a chemical saturation index as a state evaluation metric, utilizing historical operational data to guide physical mapping and optimize the dynamic response characteristics of the units. In actual operation, due to differences in location (first or last stage) and operating time, the concentration of deposited activated sludge, residual reagents, and the accumulation of reaction byproducts within each reaction unit will vary. These differences constitute the unit's chemical memory. To quantify this state, the central controller establishes and updates the chemical saturation index for each unit in real time. This index is defined as the sliding window of a unit at a specific time. The cumulative integral of the chemical reaction load borne by the interior.

[0085] The specific calculation algorithm is as follows: For any unit, record the concentration of pollutants in its influent in real time. and real-time dosage Therefore, the chemical saturation index The calculation formula is:

[0086] ;

[0087] In the formula, The set historical traceability time window length should be selected to cover the longest complete operating cycle of the system. This is the integral variable, representing a past point in time; This is a time-weighted function used to assign higher weights to recent data. It typically adopts an exponential decay form to reflect the dominant influence of the recent state of the cell on the current performance. For unit exist Real-time dosage at all times; For unit exist The concentration of pollutants in the influent at any given time; This is a time-weighted function.

[0088] From a physical perspective, the product term Directly reflecting the unit in The instantaneous reaction intensity at any given moment. The high concentration of influent combined with a large dosage indicates that the unit undergoes a vigorous chemical reaction process, potentially maintaining a high concentration of active substances, accompanied by a high rate of reaction byproduct formation. Therefore, Higher numerical values ​​indicate that the unit has been operating under high load and strong reaction conditions recently, possessing strong shock resistance and a fast reaction start-up speed; while Units with lower values ​​indicate that they have recently been under low load, in fine processing, or idle. Their internal environment is relatively clean with less residue, making them more suitable for deep purification processes. This quantitative indicator can provide a data basis for subsequent differentiated allocation based on unit characteristics.

[0089] Based on the calculated chemical saturation index, the central controller can perform the mapping and allocation of physical units to virtual link logical locations to achieve optimal configuration of the wastewater treatment system performance. Specifically, firstly, a sorting operation is performed on all available units. The central controller then reads the real-time chemical saturation index of each available unit. The cells are then sorted in descending order based on this numerical value, generating an ordered queue. In this queue, cells at the top have higher historical chemical load records, indicating an active internal reaction environment; cells at the bottom have lower saturation indices, indicating a relatively clean internal environment. Based on this queue, the central controller can execute a hierarchical mapping strategy of first-stage shock resistance and final-stage cleanliness maintenance. And for the planned... A series of parallel virtual links, the central controller first selects the top-ordered links from the queue. The units are divided into units and mapped to the first-level (i.e., head-level) position of each link. This allocation is based on the fact that the head-level unit directly withstands the high concentration shock of raw water and needs to possess strong reaction tolerance and rapid reaction kinetics. Utilizing high... The unit's chemical inertia allows for a rapid response to changes in influent load, preventing initial breakthrough effects caused by cold starts. Subsequently, the controller selects the last element in the queue. The unit is mapped to the last stage (i.e., the final stage) of each link. The main function of the final stage unit is to control the effluent quality and remove residual trace pollutants. Low-temperature effluent is selected. The unit, with its low-residue and high-cleanliness internal environment, can prevent secondary pollution caused by the unit's own release effects, ensuring that the final effluent indicators consistently meet the standards.

[0090] For the remaining units in the queue, the central controller can sequentially fill them into the intermediate positions of each link. Therefore, through this non-random allocation based on historical states, the present invention can maximize the preservation and utilization of the existing chemical properties of the units during each topology reconfiguration, ensuring a high degree of match between the current state of the physical hardware and the functional requirements of the logical positions. This avoids oscillations that may be caused by frequent switching of connection structures, thereby shortening the hydraulic and biochemical stabilization time after reconfiguration.

[0091] After completing the topology reconstruction and mapping of the physical units, this invention can execute differentiated gradient coordinated dosing strategies for reaction units at different stages. The unit at the first stage of each processing link bears the primary task of degrading the high concentration load of the influent; therefore, its control accuracy directly determines the processing pressure of subsequent stages. To this end, this embodiment adopts an enhanced feedforward control strategy for the first-stage unit. Specifically, the central controller calls the model prediction deviation signal generated by the aforementioned Smith predictor. As a correction factor. And this deviation signal It carries the error information of the current influent reacting in the pilot unit relative to the theoretical model. This error will change due to microscopic changes in water quality components, temperature fluctuations, or changes in reagent activity. Furthermore, since the pilot unit and the first-stage unit of the treatment chain are isomorphic and treat the same influent, this deviation signal is introduced into the first-stage control loop, which is equivalent to introducing a real-time correction term for the current reaction environment.

[0092] Therefore, based on this logic, the real-time dosage of the primary unit... Determined by both the basic load term and the deviation correction term, its calculation formula can be expressed as follows:

[0093] ;

[0094] In the formula, The preset basic dosage coefficient is determined based on the chemical reaction stoichiometry and the content of the active ingredient in the reagent, and characterizes the standard reagent dosage required per unit of pollutant load. This represents the current operating flow rate of the primary unit; This indicates the real-time concentration of pollutants entering the system.

[0095] Specifically, the correction term in the formula contains two key parameters: The model bias correction gain is a dimensionless positive real number used to adjust the sensitivity of the model to error control. It is usually set according to the stability requirements of the wastewater treatment system. The target effluent concentration limit is used here as a normalization factor to eliminate the influence of dimensions and establish the relative weight of the deviation.

[0096] The physical mechanism of this control algorithm lies in: when When the actual effluent concentration of the pilot unit is higher than the model's expectation, it indicates that the current water quality is difficult to treat or that there are interfering factors that inhibit the reaction. In this case, the correction term is greater than 1 and is automatically increased. Excessive dosage is applied to suppress the pollution load before it enters subsequent stages; conversely, when... When the actual reaction effect is better than expected, the correction term is less than 1, thus allowing for an appropriate reduction in the dosage, avoiding reagent waste while ensuring treatment effectiveness. Therefore, in this way, the first-stage unit achieves real-time compensation for uncertainty in reaction kinetics.

[0097] Furthermore, following the enhanced feedforward control of the aforementioned primary unit, for the reaction unit located in the middle of the virtual processing link, i.e., the link number is... (in In this invention, a proportional transfer control strategy is further employed to maintain the continuity of the reaction gradient. Specifically, due to the purification effect of the first-stage unit, the concentration of pollutants in the wastewater exhibits a non-linear decreasing trend in subsequent stages. If intermediate stages maintain the same dosage as the first-stage unit, it will lead to reagent waste and secondary pollution due to over-flushing of the final effluent pH. Therefore, the central controller needs to set the dosing command for the current stage unit based on the real-time dosage of the previous stage unit. Specifically, the first... Dosage of the unit Set as the first Level unit dosage The attenuation function is calculated according to the following proportional transfer formula:

[0098] ;

[0099] In the formula, The interstage transmission attenuation coefficient has a value range of (0, 1), and this coefficient... It is not a fixed constant, but rather dynamically corrected by the central controller based on the current reaction order characteristics of the pollutants. Specifically, when the reaction rate constant... A higher level indicates a rapid reaction and that more pollutants have been removed in the preceding stages. In this case, the central controller should be lowered. The value should be adjusted to reduce the dosage; conversely, the value should be adjusted to increase it. This value is used to maintain the driving force for subsequent reactions. Therefore, through cascading transmission, the present invention can construct a reagent concentration gradient in physical space that matches the pollutant concentration gradient.

[0100] For the last unit of the link, i.e., the sequence number is In this invention, a complementary reverse feedback control strategy is implemented to achieve effluent quality detection and maximize the utilization of remaining reagents. Specifically, a reconfigurable valve matrix is ​​driven by a central controller to construct a reverse flow path connecting the outlet of the final-stage unit to the inlet mixing zone at the beginning of the system. The central controller adjusts the return pump or regulating valve of the final-stage unit to set the return flow rate. This amount needs to be less than the influent flow rate of the primary unit. For example, the recirculation flow rate accounts for 10% to 30% of the influent flow rate. In one embodiment, the dosage of chemicals in the final stage unit... Instead of relying on feedforward calculations, a PID closed-loop feedback algorithm with a deduction term for reflux reagents is adopted, and a safety control target value lower than the emission standard is set. And calculate the measured concentration of the final effluent. Deviation from target value The central controller then adjusts the settings based on this deviation. Perform proportional-integral-differential (PID) calculations, and subtract the equivalent amount of reagent saved due to the reflux effect. Therefore, the final dosage calculation formula is as follows:

[0101] ;

[0102] In the formula, , , These are the proportional, integral, and differential gain coefficients, respectively. This represents the current reverse flow traffic; The influent flow rate of the primary unit; This refers to the current average dosage level of the entire wastewater treatment system, that is, all units. The conversion factor for the activity of refluxed reagents ( ); Represents the time from system startup time 0 to the current time. Any historical point in time between is used to calculate the cumulative amount of deviation over time.

[0103] Furthermore, the PID control ensures that the effluent quality can be forcibly brought back to the safe target under any operating conditions. Nearby, to achieve minimum control; while the subtraction item This represents a quantified deduction of the remaining reagent value in the return fluid. Furthermore, since the return wastewater may also carry unconsumed reagents and alkalinity, this portion of the return flow effectively reduces the total demand for fresh reagents. Therefore, by deducting this equivalent amount from the final stage dosing algorithm, it is possible to ensure that the final effluent meets standards while avoiding excessive dosing at the end due to the cumulative effect of return flow, thereby achieving closed-loop utilization of chemical reagents throughout the entire wastewater treatment system.

Claims

1. A modular intelligent mine wastewater treatment system, characterized in that, include: The data acquisition module is used to digitally map each wastewater treatment tank in the mine wastewater treatment system, constructing the wastewater treatment tank as a set of reaction units. The central controller sets one of the reaction units as a pilot unit for monitoring and collects the influent and effluent pollutant concentrations of all units with time synchronization in real time. The calculation module is used to execute the Smith prediction compensation algorithm based on the influent pollutant concentration and the effluent pollutant concentration to construct an instant mapping from monitoring data to the actual reaction state and calculate the real-time apparent reaction rate constant with zero delay characteristics. The decision control module is used to reverse calculate the physical processing capacity required by the unit based on the real-time apparent reaction rate constant and the preset target effluent concentration, determine the virtual cluster topology, and generate corresponding hardware driver instructions to drive the reconfigurable valve matrix to operate. The configuration module is used to establish and update the chemical saturation index of each unit in the unit set in real time, and to perform the mapping and allocation of physical units to logical positions in the virtual cluster topology based on the index. The dosing execution module is used to execute differentiated gradient coordinated dosing strategies for reaction units at different levels in the virtual cluster topology.

2. The modular intelligent mine wastewater treatment system according to claim 1, characterized in that, The calculation module executes the Smith prediction compensation algorithm specifically by: constructing a nominal model without pure lag; calculating a lag-free prediction output based on the influent pollutant concentration collected at the current moment; using a model with lag characteristics, calculating a lag prediction output based on the influent concentration data before the hydraulic retention time; performing a difference operation between the actual effluent concentration monitoring value and the lag prediction output to obtain the model prediction deviation; superimposing the model prediction deviation onto the lag-free prediction output to generate a synthetic feedback signal; and using the synthetic feedback signal to perform reverse calculation to obtain the real-time apparent reaction rate constant.

3. The modular intelligent mine wastewater treatment system according to claim 2, characterized in that, The decision control module determines the virtual cluster topology by: calculating the theoretical total reaction time required to degrade the current influent pollutant concentration to the target effluent concentration based on the real-time apparent reaction rate constant, and introducing an engineering safety factor; dividing the theoretical total reaction time by the hydraulic residence time of a single unit and rounding up to obtain the minimum number of cascade stages required for a single processing link; and dividing the total number of actually available units in the system by the minimum number of cascade stages and rounding down to obtain the maximum number of parallel chains.

4. The modular intelligent mine wastewater treatment system according to claim 2, characterized in that, The configuration module constructs the chemical saturation index by defining the index as the cumulative integral of the chemical reaction load borne by the unit within a specific time sliding window; and by recording the influent pollutant concentration and real-time dosage of the unit in real time, and then integrating the product of the two after time weighting.

5. The modular intelligent mine wastewater treatment system according to claim 4, characterized in that, The configuration module performs the mapping allocation by: sorting all available units in descending order according to their chemical saturation index; selecting the units with higher rankings and mapping them to the first-level position of the virtual cluster topology to utilize their chemical inertia to respond to the influent load; selecting the units with lower rankings and mapping them to the last-level position of the virtual cluster topology to utilize their low residual environment to prevent secondary pollution; and filling the remaining units into the intermediate-level positions.

6. The modular intelligent mine wastewater treatment system according to claim 1, characterized in that, The dosing execution module executes an enhanced feedforward control strategy for the units mapped to the primary position. Specifically, it includes: calling the model prediction deviation generated by the calculation module as a correction factor; calculating the basic load term based on the basic dosing coefficient, the current operating flow rate, and the real-time influent pollutant concentration of the system; and using the ratio of the model prediction deviation to the target effluent concentration to correct the basic load term in real time.

7. The modular intelligent mine wastewater treatment system according to claim 1, characterized in that, The dosing execution module executes a proportional transfer control strategy for units mapped to intermediate levels, specifically including: setting the dosage of the current level unit as a decay function of the dosage of the previous level unit; dynamically correcting the inter-level transfer decay coefficient based on the real-time apparent reaction rate constant, and constructing a reagent concentration gradient that matches the pollutant concentration gradient.

8. The modular intelligent mine wastewater treatment system according to claim 1, characterized in that, The dosing execution module executes a complementary reverse feedback control strategy for the unit mapped to the final stage position. Specifically, it includes: driving a reconfigurable valve matrix to construct a reverse flow path from the outlet of the final stage unit to the mixing zone of the inlet at the beginning of the system; and using a PID closed-loop feedback algorithm that includes a deduction term for backflow reagent, performing proportional-integral-differential calculations based on the deviation between the measured concentration of the final stage effluent and the safety control target value, and subtracting the equivalent value of the backflow reagent determined by the reverse backflow flow rate, the inlet flow rate of the first stage unit, and the average dosing level of the system.

9. The modular intelligent mine wastewater treatment system according to claim 1, characterized in that, When setting up the pilot unit, the data acquisition module sets the constant reference flow rate of the pilot unit to the optimal design operating flow rate of a single unit to eliminate the interference of flow fluctuations on the measurement of kinetic parameters; at the same time, the dosing unit of the pilot unit is configured to flow ratio following mode to ensure that the constant ratio of the dosage of the reagent to the influent flow rate is maintained.

10. The modular intelligent mine wastewater treatment system according to claim 1, characterized in that, The hardware driver instructions generated by the decision control module are used to control the reconfigurable valve matrix, specifically including: switching the water outlet valves of adjacent units in the virtual cluster topology to the water inlet of the next unit to form an inter-stage series path; connecting the water inlet valves of the first unit of each link to the system's main water inlet distribution pipe; and controlling the lateral interconnection valves between different parallel links to close, ensuring that each parallel link is hydraulically independent.

Images