Multiple lines concurrent processing method for call center system

Abstract

The present invention discloses a method for concurrently processing multiple lines in a call center system, which relates to the technical field of resource optimization and scheduling of call center systems. The method includes establishing a call concurrency model based on line capacity data and historical call record data, calculating call load indicators for each time period; dividing call center lines into dynamic line groups and fixed line groups based on the call load indicators; when the call load indicator is greater than a preset threshold, allocating idle lines from the fixed line group to the dynamic line group, and performing load balancing on the lines in the dynamic line group; when the call load indicator is less than or equal to the preset threshold, recycling idle lines in the dynamic line group to the fixed line group, and recording line allocation data for updating the call concurrency model, thereby achieving concurrent processing and optimized scheduling of multiple lines. The present invention solves the problem of unreasonable resource allocation in traditional call centers during business fluctuations, and also improves overall service quality through dynamic resource scheduling.

Description

Technical Field

[0001] The present invention relates to the technical field of call center system resource optimization and scheduling, and in particular to a method for concurrently processing multiple lines in a call center system. Background Art

[0002] Traditional call center systems rely primarily on manual scheduling and static resource allocation mechanisms, which can lead to performance bottlenecks under high load, such as blocked lines, delayed responses, and reduced service quality. In recent years, with the rapid development of information and communications technology, call center systems have gradually evolved towards intelligent and automated systems. In particular, scheduling systems based on big data and artificial intelligence have become essential tools for optimizing resource allocation and improving service efficiency. By incorporating real-time monitoring, data analysis, and artificial intelligence algorithms, call center systems can perceive the operating status of each communication line in real time and dynamically adjust resource allocation based on actual load conditions, significantly improving the system's load handling capacity and resource utilization efficiency.

[0003] However, existing technologies still have certain shortcomings in concurrent processing and optimized scheduling. Traditional load forecasting models mostly rely on simple linear regression or rule-based scheduling algorithms. These methods are difficult to deal with the complex multi-dimensional resource scheduling problems in call centers. Although existing systems can perform a certain degree of resource scheduling through simple load monitoring methods, they usually ignore the impact of multiple factors such as line quality and response delay on resource allocation, resulting in inaccurate scheduling decisions and waste of resources. At the same time, existing load forecasting models fail to fully consider the dynamic change characteristics in time series data, lack sufficient prediction accuracy, and cannot cope with load fluctuations during peak hours. In addition, traditional methods usually only consider a single scheduling objective when scheduling resources, such as minimizing response delay or maximizing system throughput. They lack comprehensiveness and flexibility, and it is difficult to achieve balanced optimization of multiple objectives.

[0004] Therefore, existing technologies face problems such as uneven resource allocation, long response delays, and insufficient system processing capacity when handling concurrent requests from multiple lines. This makes it impossible to effectively guarantee stable call center operation and service quality under high load conditions. To address these issues, the present invention proposes a method for concurrently processing multiple lines in a call center system. This method achieves efficient concurrent processing and dynamic optimized scheduling of multiple lines, addressing the shortcomings of existing technologies in complex load forecasting and multi-objective scheduling optimization. It can more accurately allocate resources and improve the overall performance and response speed of the system. Summary of the Invention

[0005] In view of the problems of uneven load distribution, long response delay and resource waste in the existing call center system during the concurrent processing of multiple lines, the present invention is proposed.

[0006] Therefore, the problem to be solved by the present invention is how to dynamically adjust the resource allocation of call center lines through load forecasting and multi-dimensional resource allocation optimization, realize efficient concurrent processing of multiple lines, improve the overall performance and response speed of the system, and thus effectively deal with resource bottlenecks and performance degradation problems under high concurrency and high load conditions.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] In a first aspect, an embodiment of the present invention provides a method for concurrently processing multiple lines in a call center system, comprising: establishing a call concurrency model based on line capacity data and historical call record data, and calculating call load indicators for each time period; dividing the call center lines into dynamic line groups and fixed line groups based on the call load indicators; when the call load indicators are greater than a preset threshold, allocating idle lines from the fixed line group to the dynamic line group, and performing load balancing on the lines of the dynamic line group; when the call load indicators are less than or equal to the preset threshold, reclaiming idle lines in the dynamic line group to the fixed line group, and recording line allocation data for updating the call concurrency model, thereby realizing concurrent processing and optimized scheduling of multiple lines.

[0009] As a preferred solution of the method for concurrently processing multiple lines in a call center system described in the present invention, the method for calculating the call load index is as follows: collecting line capacity data of the call center system and obtaining historical call record data, wherein the line capacity data includes the maximum concurrent number, peak load capacity and available duration of each communication line; establishing a capacity constraint matrix based on the line capacity data, and performing time series analysis on the historical call record data, and extracting call pattern characteristics using the exponential smoothing method; based on the capacity constraint matrix and the call pattern characteristics, using the support vector regression algorithm to construct a call concurrency model and calculate the call load index.

[0010] As a preferred embodiment of the method for concurrently processing multiple lines in a call center system according to the present invention, the historical call record data includes call time distribution, call duration distribution, and call density distribution; the capacity constraint matrix is ​​divided into 48 time segments based on a 24-hour period; the call concurrency model uses a Markov chain to describe call transfer probability; and the specific formula for the call load indicator is as follows:

[0011]

[0012] Where L(t) is the call load index at time t, ω i is the weight coefficient of the i-th line, λ is the time attenuation factor, t iis the activation time of the i-th line, C i is the maximum number of concurrent connections on the i-th line, Δ is the time window size, P ij is the call transfer probability matrix element, H j (τ) is the historical call density function, k is the number of observation time segments, D s is the call density deviation in the sth time segment, m is the total number of lines in the system, and n is the number of call states.

[0013] As a preferred embodiment of the method for concurrently processing multiple lines in a call center system according to the present invention, the method further comprises: dividing the call center lines into dynamic line groups and fixed line groups based on the call load index, including: setting line grouping thresholds based on the fluctuation pattern of the call load index, and analyzing the call load index of each time segment, wherein the line grouping thresholds include a load reference value, an upper fluctuation limit value, and a lower fluctuation limit value; classifying time segments in which the call load index is greater than the load reference value and the fluctuation amplitude exceeds 30% as a dynamic control interval; and classifying time segments in which the call load index is less than or equal to the load reference value and the fluctuation amplitude is within 15% as a fixed guarantee interval; setting dynamic line groups and fixed line groups based on the dynamic control interval and the fixed guarantee interval; and establishing a coordination mechanism between the dynamic line groups and the fixed line groups. When the dynamic line group has insufficient capacity, idle resources of the fixed line group can be temporarily called upon, and a line group switching buffer is set, wherein the line group switching buffer is used to smooth load fluctuations during the switching process.

[0014] As a preferred embodiment of the method for concurrently processing multiple lines in a call center system according to the present invention, when the call load indicator exceeds a preset threshold, idle lines are allocated from the fixed line group to the dynamic line group, and load balancing is performed on the lines of the dynamic line group. The method includes: monitoring the changing trend of the call load indicator. When it is detected that the call load indicator exceeds the preset threshold for three consecutive sampling periods, a line allocation mechanism is triggered to calculate the real-time load rate of the dynamic line group; if the real-time load rate exceeds 85% and still shows an increasing trend, a cross-group line allocation procedure is initiated; the idle line status in the fixed line group is detected, and lines that have been idle for an expected period of time and have good performance indicators are selected as resources to be allocated; line allocation is performed, and the number of lines allocated is determined based on the load excess ratio, and a single allocation does not exceed 20% of the total capacity of the fixed line group; performance warm-up is performed on the newly allocated lines, call tasks are allocated to the new lines through a gradual loading method, and a load balancing algorithm is simultaneously executed on all lines of the dynamic line group.

[0015] As a preferred embodiment of the method for concurrently processing multiple lines in a call center system according to the present invention, the line allocation mechanism includes the following steps: collecting the numerical changes of the call load indicator and calculating the real-time load rate of the dynamic line group, wherein the real-time load rate is calculated by the ratio of the number of active calls in the current period to the total number of available lines; performing trend analysis on the real-time load rate, and if the real-time load rate exceeds 85% and shows an upward trend, automatically detecting the idle line status in the fixed line group, screening out continuously idle and high-performance lines as allocation resources, and determining the specific number of lines to be allocated based on the current load excess ratio; preheating the performance of the selected new lines, and allocating call tasks in a progressive manner, while executing a load balancing algorithm based on the minimum number of connections principle on all lines in the dynamic line group to achieve optimal resource allocation.

[0016] As a preferred solution of the method for concurrently processing multiple lines in a call center system according to the present invention, when the call load index is less than or equal to the preset threshold, the idle lines in the dynamic line group are recycled to the fixed line group, and the line allocation data is recorded for updating the call concurrency model to achieve concurrent processing and optimized scheduling of multiple lines, including: monitoring the changes in the call load index, triggering the line recycling mechanism when it is detected that the call load index is lower than the preset threshold for 5 consecutive sampling periods, and analyzing the resource occupancy of each line in the dynamic line group to identify the call volume that is continuously lower than 4 within the expected time. 0% of the lines are selected as objects to be recovered; line recovery preprocessing is performed to orderly migrate existing call tasks on the lines to be recovered; line recovery operations are implemented in batches, and the number of lines recovered in each batch does not exceed 15% of the current capacity of the dynamic line group, and an observation period is set between adjacent batches; the recovered lines are reintegrated into the fixed line group, and status reset and performance testing are performed, and detailed data of this line recovery is recorded, wherein the detailed data includes the recovery time point, the number of recovered lines, load change trends and service quality indicators; the detailed data is input into the call concurrency model for online learning to optimize the model's prediction ability for load changes.

[0017] In a second aspect, an embodiment of the present invention provides a system for concurrently processing multiple lines in a call center system, comprising: a call concurrency model construction module, configured to establish a call concurrency model based on line capacity data and historical call record data, and calculate call load indicators for each time period; a division module, configured to divide the call center lines into dynamic line groups and fixed line groups based on the call load indicators; a load allocation and balancing module, configured to allocate idle lines from the fixed line group to the dynamic line group when the call load indicator is greater than a preset threshold, and to perform load balancing on the lines of the dynamic line group; a line recovery and model update module, configured to recover idle lines in the dynamic line group to the fixed line group when the call load indicator is less than or equal to the preset threshold, and to record line allocation data for updating the call concurrency model, thereby realizing concurrent processing and optimized scheduling of multiple lines.

[0018] In a third aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the method for concurrently processing multiple lines in a call center system as described in the first aspect of the present invention are implemented.

[0019] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of the method for concurrently processing multiple lines in a call center system as described in the first aspect of the present invention are implemented.

[0020] The beneficial effects of the present invention are: by establishing a call concurrency model based on historical data, accurate prediction of business load is achieved, and a dynamic and static combined line group division mechanism is adopted to provide a reliable basis for resource allocation. The intelligent allocation strategy and load balancing algorithm triggered by thresholds are combined to ensure efficient utilization of system resources. At the same time, continuous optimization of the model is achieved through timely resource recovery and data feedback mechanisms. This method not only solves the problem of unreasonable resource allocation in traditional call centers during business fluctuations, but also improves the overall service quality through dynamic resource scheduling, ultimately achieving the technical effects of improving line resource utilization, reducing operating costs and enhancing system reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them:

[0022] Figure 1This is a flowchart of a method for concurrently processing multiple lines in a call center system according to Example 1. DETAILED DESCRIPTION

[0023] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0024] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0025] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0026] Example 1

[0027] Reference Figure 1 , which is the first embodiment of the present invention, provides a method for concurrently processing multiple lines in a call center system, comprising:

[0028] S1: Establish a call concurrency model based on line capacity data and historical call record data, and calculate the call load index for each time period.

[0029] Specifically, the call load index is calculated by collecting line capacity data from the call center system and obtaining historical call record data; establishing a capacity constraint matrix based on the line capacity data, performing time series analysis on the historical call record data, and using exponential smoothing to extract call pattern characteristics;

[0030] It should be noted that line capacity data includes the maximum number of concurrent users, peak load capacity, and available time of each communication line; historical call record data includes call time distribution, call duration distribution, and call density distribution; the capacity constraint matrix is ​​divided into 48 time segments based on a 24-hour cycle; and the call concurrency model uses a Markov chain to describe call transfer probability.

[0031] Furthermore, based on the capacity constraint matrix and call pattern characteristics, a support vector regression algorithm is used to construct a call concurrency model and calculate the call load index.

[0032] Furthermore, the specific formula for the call load indicator is as follows:

[0033]

[0034] Where L(t) is the call load index at time t, ω i is the weight coefficient of the i-th line, λ is the time attenuation factor, t i is the activation time of the i-th line, C i is the maximum number of concurrent connections on the i-th line, Δ is the time window size, P ij is the call transfer probability matrix element, H j (τ) is the historical call density function, k is the number of observation time segments, D s is the call density deviation in the sth time segment, m is the total number of lines in the system, and n is the number of call states.

[0035] S2: Based on the call load indicator, divide the call center lines into dynamic line groups and fixed line groups.

[0036] Specifically, based on the fluctuation pattern of call load indicators, line grouping thresholds are set and the call load indicators of each time segment are analyzed; time segments where the call load indicators are greater than the load reference value and the fluctuation range exceeds 30% are classified as dynamic control intervals; at the same time, time segments where the call load indicators are less than or equal to the load reference value and the fluctuation range is within 15% are classified as fixed protection intervals.

[0037] Furthermore, based on the dynamic control interval and the fixed guarantee interval, dynamic line groups and fixed line groups are set up; a coordination mechanism between the dynamic line groups and the fixed line groups is established. When the dynamic line group has insufficient capacity, the idle resources of the fixed line group can be temporarily called, and a line group switching buffer is set up, where the line group switching buffer is used to smooth load fluctuations during the switching process.

[0038] It should be noted that the dynamic line group adjusts the number of lines as the call load index fluctuates; the fixed line group maintains a basic line scale of no less than 40% of the total number of lines; the line grouping threshold includes the load baseline value, the upper limit of fluctuation and the lower limit of fluctuation.

[0039] S3: When the call load indicator is greater than a preset threshold, allocate idle lines from the fixed line group to the dynamic line group, and perform load balancing on the lines of the dynamic line group.

[0040] Specifically, the changing trend of the call load indicator is monitored. When it is detected that the call load indicator exceeds the preset threshold for three consecutive sampling periods, the line allocation mechanism is triggered to calculate the real-time load rate of the dynamic line group; if the real-time load rate exceeds 85% and still has an upward trend, the cross-group line allocation program is started.

[0041] It should be noted that the preset threshold is a dynamic threshold range that balances system stability and resource utilization by analyzing the maximum concurrency during peak hours, calculating the call connection rate per unit time, evaluating the performance overhead of line switching, and combining the fluctuation trend of business volume in historical data.

[0042] Furthermore, the line allocation mechanism includes the following steps: collecting the numerical changes of call load indicators, calculating the real-time load rate of the dynamic line group, where the real-time load rate is calculated by the ratio of the number of active calls in the current period to the total number of available lines; performing trend analysis on the real-time load rate, if the real-time load rate exceeds 85% and shows an upward trend, automatically detecting the idle line status in the fixed line group, screening out continuously idle and high-performance lines as allocation resources, and determining the specific number of lines to be allocated based on the current load excess ratio; preheating the performance of the selected new lines, and allocating call tasks in a progressive manner, while executing a load balancing algorithm based on the minimum number of connections principle on all lines in the dynamic line group to achieve optimal resource allocation.

[0043] Furthermore, the idle line status in the fixed line group is detected, and lines that are continuously idle within the expected time and have good performance indicators are screened as resources to be allocated; line allocation operations are implemented, and the number of lines allocated is determined based on the load excess ratio, and a single allocation does not exceed 20% of the total capacity of the fixed line group; the performance of the newly allocated lines is preheated, and call tasks are allocated to the new lines through gradual loading. At the same time, a load balancing algorithm is executed on all lines in the dynamic line group.

[0044] S4: When the call load indicator is less than or equal to the preset threshold, the idle lines in the dynamic line group are recycled to the fixed line group, and line allocation data is recorded to update the call concurrency model to achieve concurrent processing and optimized scheduling of multiple lines.

[0045] Specifically, the changes in call load indicators are monitored. When it is detected that the call load indicator is lower than the preset threshold for five consecutive sampling periods, the line recovery mechanism is triggered, and the resource occupancy of each line in the dynamic line group is analyzed. Lines whose call volume is continuously lower than 40% within the expected time are identified as lines to be recovered.

[0046] Furthermore, line recovery preprocessing is performed to orderly migrate existing call tasks on the lines to be recovered; line recovery operations are implemented in batches, and the number of lines recovered in each batch does not exceed 15% of the current capacity of the dynamic line group, and an observation period is set between adjacent batches.

[0047] Furthermore, the recovered lines are reintegrated into the fixed line group, status reset and performance testing are performed, and detailed data of this line recovery is recorded; the detailed data is input into the call concurrency model for online learning to optimize the model's ability to predict load changes.

[0048] It should be noted that the detailed data includes the recovery time point, the number of recovery lines, the load change trend and the service quality indicators.

[0049] In summary, the present invention achieves accurate prediction of business load by establishing a call concurrency model based on historical data, adopts a dynamic and static combined line group division mechanism to provide a reliable basis for resource allocation, and combines threshold-triggered intelligent allocation strategies and load balancing algorithms to ensure efficient utilization of system resources. At the same time, through timely resource recovery and data feedback mechanisms, continuous optimization of the model is achieved. This method not only solves the problem of unreasonable resource allocation in traditional call centers during business fluctuations, but also improves the overall service quality through dynamic resource scheduling, ultimately achieving the technical effects of improving line resource utilization, reducing operating costs and enhancing system reliability.

[0050] Furthermore, this embodiment also provides a system for concurrently processing multiple lines in a call center system, including: an acquisition module for acquiring real-time signals from multiple communication lines in a call center, fusing the real-time signals and converting them into original line monitoring data, wherein the real-time signals include call load signals, line quality signals, and response delay signals; a feature extraction module for performing wavelet transform on the original line monitoring data, extracting its high-frequency components and low-frequency components, and constructing a line performance feature vector using the high-frequency components and the low-frequency components; a load index prediction module for inputting the line performance feature vector into a pre-trained long-short-term memory neural network model to obtain a load index of the call center line, wherein the load index represents the allocation and processing capacity of line resources; a resource allocation and scheduling optimization module for establishing a multidimensional line resource allocation model based on the load index, and calculating the resource weight parameters and scheduling strategy parameters of the multidimensional line resource allocation model through the maximum likelihood estimation method to achieve concurrent processing and optimized scheduling of multiple lines.

[0051] This embodiment also provides a computer device suitable for the case of a method for concurrently processing multiple lines in a call center system, comprising a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for concurrently processing multiple lines in a call center system as proposed in the above embodiment.

[0052] The computer device may be a terminal, comprising a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device comprises a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner may be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.

[0053] Example 2

[0054] Referring to Table 1, which is a second embodiment of the present invention, this embodiment provides a method for concurrently processing multiple lines in a call center system. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0055] Specifically, we collected six months of historical call data from a provincial customer service center, including call volume, call duration, and call density by time of day. We used Python data analysis tools to preprocess the historical data, using the pandas library for time series analysis, and the sklearn library to implement the support vector regression algorithm and build a concurrent model.

[0056] Furthermore, during the model training phase, historical data was divided into 48 time segments, each 30 minutes long, based on a 24-hour cycle. Exponential smoothing was used to extract call pattern characteristics and establish a capacity constraint matrix. This capacity constraint matrix takes into account parameters such as the maximum number of concurrent calls per communication line (ranging from 80 to 120 concurrent calls), peak load capacity (up to 90% load during peak hours), and available time (normal operation > 23 hours / day). A Markov chain was used to describe call transfer probabilities, with a time decay factor set to 0.85 and an observation window size of 12 time segments.

[0057] Furthermore, based on the trained model, the 300 lines were initially divided into 200 fixed lines (fixed line group) and 100 dynamic lines (dynamic line group). The load baseline value in the line grouping thresholds was set to 65%, the upper fluctuation limit was 85%, and the lower fluctuation limit was 40%. When the system detects that the load index exceeds 85% for three consecutive sampling periods (each sampling period is 5 minutes), the line allocation mechanism is triggered; when it falls below 40% for five consecutive sampling periods, the line recycling mechanism is activated.

[0058] Specifically, as shown in Table 1, the dynamic resource allocation method proposed in the present invention has significant advantages over the traditional fixed allocation scheme. In terms of service quality, the average connection rate has increased from 82.5% to over 95%, and the average response time has been reduced from 45.8 seconds to less than 10 seconds. This is due to the system's ability to quickly adjust resource allocation according to real-time load indicators, effectively responding to business peak demand.

[0059] Table 1 Comparison of performance indicators of call center dynamic resource allocation system

[0060]

[0061] Furthermore, in terms of resource utilization efficiency, through the dynamic allocation mechanism, the overall system resource utilization rate has increased from 58.6% to over 85%. In particular, on weekdays, through the intelligent prediction allocation solution, the resource utilization rate can reach as high as 91.2%. At the same time, the system's peak concurrent processing capacity has been significantly improved, from the original 850 concurrent connections to 1350 concurrent connections, an increase of 58.8%. In terms of system stability, by introducing a line group switching buffer and a progressive load distribution mechanism, the system stability index has increased from 0.82 to over 0.95. In particular, after adopting the intelligent prediction allocation solution, the system stability index reached 0.99, indicating that the present invention can maintain stable operation of the system while ensuring high efficiency.

[0062] Furthermore, in terms of economic benefits, through dynamic resource allocation and intelligent forecasting mechanisms, the system's operating costs have been reduced by 15.8%-25.3% across different scenarios. This is primarily due to the system's ability to automatically recycle idle resources during off-peak periods, avoiding resource waste, while also reducing the need for additional resources during peak periods through accurate resource forecasting and allocation.

[0063] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for concurrently processing multiple lines in a call center system, characterized by: include, Establish a call concurrency model based on line capacity data and historical call record data, and calculate call load indicators for each time period; Based on the call load indicator, the call center lines are divided into dynamic line groups and fixed line groups; When the call load indicator is greater than a preset threshold, idle lines are allocated from the fixed line group to the dynamic line group, and load balancing is performed on the lines of the dynamic line group; When the call load indicator is less than or equal to the preset threshold, the idle lines in the dynamic line group are recycled to the fixed line group, and line allocation data is recorded for updating the call concurrency model to achieve concurrent processing and optimized scheduling of multiple lines; Collect line capacity data from the call center system and obtain historical call record data; Establishing a capacity constraint matrix based on the line capacity data, performing time series analysis on the historical call record data, and extracting call pattern features using an exponential smoothing method; Based on the capacity constraint matrix and the call pattern characteristics, a support vector regression algorithm is used to construct a call concurrency model and calculate a call load index; The line capacity data includes the maximum number of concurrent users, peak load capacity, and available time of each communication line; the historical call record data includes call time distribution, call duration distribution, and call density distribution; the capacity constraint matrix is ​​divided into 48 time segments based on a 24-hour cycle; the call concurrency model uses a Markov chain to describe call transfer probability; According to the fluctuation pattern of the call load index, a line grouping threshold is set, and the call load index of each time segment is analyzed, wherein the line grouping threshold includes a load reference value, an upper fluctuation limit value, and a lower fluctuation limit value; The time segment in which the call load index is greater than the load benchmark value and the fluctuation range exceeds 30% is classified as a dynamic control period; At the same time, the time segments where the call load index is less than or equal to the load benchmark value and the fluctuation range is within 15% are classified as fixed guarantee intervals; Setting a dynamic line group and a fixed line group based on the dynamic control interval and the fixed guarantee interval; A coordination mechanism is established between the dynamic line group and the fixed line group. When the dynamic line group has insufficient capacity, the idle resources of the fixed line group can be temporarily called upon, and a line group switching buffer is set, wherein the line group switching buffer is used to smooth load fluctuations during the switching process.

2. The method for concurrently processing multiple lines in a call center system according to claim 1, wherein: The specific formula of the call load indicator is as follows: ； in, is the call load index at time t, is the weight coefficient of the i-th line, is the time decay factor, is the activation time of the i-th line, is the maximum number of concurrent connections on the ith line, is the time window size, is the call transfer probability matrix element, is the historical call density function, is the number of observation time segments, is the call density deviation of the sth time segment, is the total number of lines in the system, The number of call states.

3. The method for concurrently processing multiple lines in a call center system according to claim 2, wherein: When the call load indicator is greater than a preset threshold, idle lines are allocated from the fixed line group to the dynamic line group, and load balancing is performed on the lines of the dynamic line group, including: Monitor the changing trend of the call load indicator. When it is detected that the call load indicator exceeds the preset threshold for three consecutive sampling periods, the line allocation mechanism is triggered to calculate the real-time load rate of the dynamic line group. If the real-time load rate exceeds 85% and still shows an upward trend, the cross-group line allocation process is initiated. Detecting the idle line status in the fixed line group, and selecting lines that are continuously idle within an expected time and have good performance indicators as resources to be allocated; Implement line allocation operations, and the number of lines allocated is determined based on the load excess ratio, and a single allocation does not exceed 20% of the total capacity of the fixed line group; The performance of the newly deployed lines is preheated, call tasks are allocated to the new lines through gradual loading, and the load balancing algorithm is executed on all lines of the dynamic line group.

4. The method for concurrently processing multiple lines in a call center system according to claim 3, wherein: The line allocation mechanism includes the following steps: collecting numerical changes of the call load indicator and calculating a real-time load rate of the dynamic line group, wherein the real-time load rate is calculated by the ratio of the number of active calls in the current period to the total number of available lines; Performing a trend analysis on the real-time load rate; if the real-time load rate exceeds 85% and shows an upward trend, automatically detecting the idle line status in the fixed line group, selecting lines that are continuously idle and have good performance as deployment resources, and determining the specific number of lines to be deployed based on the current load excess ratio; The performance of the selected new lines is preheated, and call tasks are allocated in a progressive manner. At the same time, a load balancing algorithm based on the principle of minimum number of connections is executed on all lines in the dynamic line group to achieve optimal resource allocation.

5. The method for concurrently processing multiple lines in a call center system according to claim 4, wherein: When the call load indicator is less than or equal to the preset threshold, the idle lines in the dynamic line group are recycled to the fixed line group, and line allocation data is recorded for updating the call concurrency model to achieve concurrent processing and optimized scheduling of multiple lines, including: monitoring changes in the call load indicator, and triggering a line recycling mechanism when it is detected that the call load indicator is below a preset threshold for five consecutive sampling periods, analyzing resource usage of each line in the dynamic line group, and identifying lines whose call volume is continuously below 40% within the expected time as lines to be recycled; Perform line recovery pre-processing and orderly migrate existing call tasks on the recovered lines; Implement line recovery operations in batches, with the number of lines recovered in each batch not exceeding 15% of the current capacity of the dynamic line group, and an observation period set between adjacent batches; Reintegrate the recovered lines into the fixed line group, perform status reset and performance testing, and record detailed data of this line recovery, including recovery time, number of recovered lines, load change trend, and service quality indicators; The detailed data is input into the call concurrency model for online learning to optimize the model's ability to predict load changes.

6. A system for concurrently processing multiple lines in a call center system, based on the method for concurrently processing multiple lines in a call center system according to any one of claims 1 to 5, characterized in that: include, A call concurrency model building module is used to build a call concurrency model based on line capacity data and historical call record data, and calculate call load indicators for each time period; a division module, which divides the call center lines into dynamic line groups and fixed line groups based on the call load indicator; A load allocation and balancing module, configured to allocate idle lines from the fixed line group to the dynamic line group when the call load indicator is greater than a preset threshold, and perform load balancing on the lines of the dynamic line group; The line recovery and model update module is used to recover idle lines in the dynamic line group to the fixed line group when the call load indicator is less than or equal to the preset threshold, and record line allocation data for updating the call concurrency model to achieve concurrent processing and optimized scheduling of multiple lines.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for concurrently processing multiple lines in a call center system according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for concurrently processing multiple lines in a call center system according to any one of claims 1 to 5 are implemented.

Images